What is This Document for?

Using data collected through the Student Wellbeing and Success Initiative (SWaSI), the purposes of this report are to:

Estimate rates of food insecurity among undergraduates at the University of Oregon (UO)
Document and describe relationships between food insecurity, demographic characteristics, and wellbeing factors among UO undergraduates
Make recommendations informed by the data

Who is SWaSI?

The Student Wellbeing and Success Initiative (SWaSI) is an:

Educational program designed to improve wellbeing and success outcomes, particularly among historically marginalized populations (e.g., first-generation students, Black students) and to provide undergraduates and graduates with experiential learning opportunities doing research
Assessment and evaluation program designed to gauge whether various programmatic activities, including but not limited to the ones internal to the Initiative itself, are meeting their stated goals
Ongoing, multicohort, longitudinal research program designed to holistically understand institutional inputs to students’ wellbeing and success across their experience at the university

Better understanding better informs us, affording opportunities to improve institutional practices that support students’ learning and development and foster their achievement and persistence. Unlike many for-profit companies, which use data to sell people, this is a strategy that uses data to empower people.

The Student Wellbeing and Success Initiative is led by Brian Clark, Assistant Director of the Office of Assessment and Research in the Division of Student Life. It is supported primarily by the Office of the Vice President for Student Life with integral support from units of the Division of Student Life – the Department of Physical Education and Recreation, the Erb Memorial Union, and the Office of the Dean of Students – and across campus from many and varied units in Global Engagement, the Office of the Provost, Student Services and Enrollment Management, UO Libraries, and Undergraduate Education and Student Success.

What is Food Security?

The United States Department of Agriculture (USDA) defines food security as access to and ready availability of enough nutritionally adequate and safe food for an active, healthy life, and includes the assured ability to acquire food in socially acceptable ways (e.g., without scavenging, stealing, etc.). The UO Food Security Task Force defines food security as reliable access to healthy, nutritious, culturally appropriate food. The preceding definitions are conceptualized at an individual level or, at most, a household level. Although food security can be conceptualized at higher levels (see Pinstrup-Andersen, 2009 for a brief discussion; see also Jones et al., 2013 for a comprehensive review), we adopt an individual-level definition in the present research.

Since the first study exploring food insecurity in a college student population was published (Chaparro et al., 2009), researchers have consistently found that the prevalence of food insecurity among college students is higher than the national household rate (Abbey et al., 2022; Bruening et al., 2017; Nikolaus et al. 2020; The Hope Center for Student Basic Needs at Temple University, 2024). Three recent systematic reviews of the literature found that the average prevalence estimates of food insecurity among college students in the United States ranged from 32% to 41% (Abbey et al., 2022; Bruening et al., 2017; Nikolaus et al., 2020). In 2023-2024, a national survey administered to 74,350 students attending 91 universities in 16 states found that 41% of students experienced food insecurity (The Hope Center for Student Basic Needs at Temple University, 2024). Rates of food insecurity among students attending public institutions in Oregon are similar or even higher. Researchers have estimated the prevalence of food insecurity among students at Oregon State University to be 38.9% (Edwards et al., 2022), 47.0% at Portland State University (Townley et al., 2020), and 58.8% at Western Oregon University (Patton-Lopez et al., 2014). In contrast, the prevalence of household food insecurity in the United States was 10.2% in 2021 (Coleman-Jensen et al., 2022), 12.8% in 2022 (Rabbit et al., 2023), and 13.5% in 2023 (Rabbit et al., 2024).

There are several documented associations with food insecurity, and food insecurity is disproportionately experienced among different demographic subpopulations. Key demographic characteristics associated with or predictive of higher food insecurity include having a marginalized race or ethnic identity, (Bruening et al., 2017; DeBate et al., 2021; Edwards et al., 2022; El Zein et al., 2019; Haskett et al., 2020; Laska et al., 2021; Martinez et al., 2017; Owens et al., 2020; Tanner et al., 2023; Townley et al., 2020; The Hope Center for Student Basic Needs at Temple University, 2024) and identifying as female, non-binary, or 2SLGBTQIA3+ (Edwards et al., 2022; Haskett et al., 2020; Laska et al., 2021; Martinez et al., 2017; Olfert et al., 2021; Townley et al., 2020). Transfer students (Tanner et al., 2023) and first-generation students (Cuite et al., 2023; El Zein et al., 2019; Guzman et al., 2022; Olfert et al., 2019; Tanner et al., 2023) also have greater probability of experiencing food insecurity. Food insecurity rates are also disproportionately higher among students with disabilities or who are neurodivergent (McArthur et al., 2020; Olfert et al., 2021; Stott et al., 2023; Townley et al., 2020). Economic and financial factors associated with a greater likelihood of food insecurity include being from a lower socioeconomic background (Laska et al., 2021; Tanner et al., 2023), being a Pell Grant recipient (El Zein et al., 2019), receiving financial aid (Martinez et al., 2017, Olfert et al., 2021), and experiencing childhood food insecurity (Martinez et al., 2017; Olfert et al., 2021). Being a parent (Olfert et al., 2021; The Hope Center for Student Basic Needs at Temple University, 2024) or single parent (Owens et al., 2020) is associated with greater odds of food insecurity compared to students who are not parents. In relation to outcomes, food insecurity is associated with higher stress, higher loneliness (DeBate et al., 2021; Guzman et al., 2022), greater depression, anxiety and hopelessness (Raskind et al., 2019), higher psychological distress (Becerra et al., 2020; Guzman et al., 2022), poorer overall mental health (Becerra et al., 2020; Martinez et al., 2020), poorer physical health (Martinez et al., 2019), and poorer academic outcomes, including lower GPA, attendance, and persistence (Camelo & Elliott, 2019; Haskett et al., 2021; Mechler et al., 2024; Martinez et al., 2020; Robbins et al., 2022; Silva et al., 2017). Outside of the higher education context, research has found that higher food security is associated with lower stress and rates of mental illness (Martin et al., 2016), as well as higher life satisfaction (Graham et al., 2019) and overall psychological wellbeing (MacNeil et al., 2020).

In the context of the UO, we have found that the prevalence of food insecurity among undergraduates is comparable with recent national estimates and those observed at Oregon State University and Portland State University. We have also confirmed many of the above associations over the past three years (Clark & Delgado-Riley, 2021; 2022; 2023). The rates of food insecurity among UO undergraduates were 21.8% in 2021, 34.0% in 2022, and 39.7% in 2023 – double to triple the national household rates of 10.2% in 2021 (Coleman-Jensen et al., 2022), 12.8% in 2022 (Rabbit et al., 2023), and 13.5% in 2023 (Rabbit et al., 2024). Students of lower socioeconomic status, students with disabilities or who are neurodivergent, and 2SLGBTQIA3+, Latine/a/o/x, and transfer students were more likely to be food insecure than their peers. Additionally, being food insecure was significantly associated with higher stress, higher sadness, and lower life satisfaction, even after controlling for potentially confounding demographics.

Brief Method

This year, as with previous years, our approach to analyzing the data was a mixture of confirmatory and exploratory. We aimed to replicate and extend by:

Estimating rates of food insecurity among UO undergraduates
Focusing exclusively on U.S. undergraduates due to continued low participation among international undergraduates
Including confidence intervals to the overall estimates of food security
Investigating whether previously observed associations between food security, demographic characteristics, and wellbeing factors among UO undergraduates remained or changed
Updating demographics to:
- Include disability/neurodivergence status and a new composite variable of socioeconomic status based on parent education, self-reported global and family socioeconomic status, and high school quality
- Exclude the demographics that have not been significant predictors in the past (binary sex/gender and residency status were dropped)
Letting the data drive decisions about the use of control covariates
Including contextualizing results from other data sources:
- Estimates of increased food cost
- National College Health Assessment and Hope Center Student Basic Needs Survey estimates of food insecurity among first-through-fourth-year U.S. undergraduates at the UO
- Estimates of food insecurity at other public institutions nationally and in Oregon
Expanding and improving basic needs data analysis
Improving descriptives of demographic variables in the Supplemental Method

Aligning with the USDA definition of food security, we measured food security using USDA methods. To verify or explore relationships with demographic variables, we joined food security data with demographics from the Student Wellbeing and Success Initiative and institutional student records (primarily admissions applications data). Demographics examined were disability/neurodivergence status, gender identity/sexuality, race/ethnicity, socioeconomic status (a composite variable), and transfer status. Measurements of stress, sadness, and life satisfaction routinely collected through the Student Wellbeing and Success Initiative were used for verifying associations between food security and wellbeing constructs.

The 10-item USDA Food Security Module classifies four levels along a continuum of food security.

High Food Security: No problems, or anxiety about, consistently accessing adequate food
Marginal Food Security: Problems at times, or anxiety about, accessing adequate food, but quality, variety, and quantity of food intake not substantially reduced
Low Food Security: Reduced quality, variety, and desirability of diets, but quantity of food intake and normal eating patterns not substantially disrupted
Very Low Food Security: Eating patterns disrupted, food intake reduced due to lack of money and other resources

These categories can be further subdivided as a binary variable: Food Secure (high & marginal food security) vs. Food Insecure (low & and very low food security). This binary approach was used in the present analyses.

Food security data were collected at UO in spring 2024, 2023, 2022, and 2021. The recall time frame for the food security measurement was 12 months, which entailed a span from spring to spring (i.e., spring 2023 to spring 2024, spring 2022 to spring 2023, spring 2021 to spring 2022, and spring 2020 to spring 2021). The National College Health Assessment (spring 2024) used a short, adapted form of the USDA module and the Hope Center Student Basic Needs Survey (winter 2023) used the same 10-item form of the USDA module that we used. Both the National College Health Assessment and Hope Student Basic Needs Survey had a 30-day recall time frame.

For greater detail, see the Supplemental Method section.

Summary of Findings

Overall Estimates
- Between spring 2023 and spring 2024, 39.50% of first-through-fourth-year U.S. undergraduates attending the UO were food insecure. This is roughly three times the national household rate of 13.5% in 2023 (Rabbit et al.). Among UO undergraduates, food insecurity was triple the national household rate in 2023 (39.70%), triple the national household rate in 2022 (34.00%), and double the national household rate in 2021 (21.80%).
  - However, 2SLGBTQIA3+ students were likely oversampled this year. Because they tend to be less food secure than do other students, this appears to have inflated the overall estimated food insecurity rate by about 2 percentage points. Still, 37.5% is a rather high rate of food insecurity. In 2021, the sample did not include fourth-year students, and Black/Indigenous students were undersampled. As such, the overall rate in 2021 may have been an underestimate.
By Demographics
- Disability/neurodivergence status, socioeconomic status, gender/sexuality (2SLGBTQIA3+ vs. Not 2SLGBTQIA3+), and transfer status were the strongest demographic predictors of food insecurity.
- Race/ethnicity did NOT account for unique variation in food security once other variables – importantly, disability/neurodivergence and socioeconomic status – were taken into account.
Wellbeing Consequences
- As expected, being food insecure remains significantly associated with higher stress, higher sadness, and lower life satisfaction, even after controlling for potentially confounding demographics.
Student Voices – Basic Needs Support
- In spring 2022, we introduced a general open-ended question: How can the UO better support your basic needs (food, housing, safety, belonging, self-esteem, fulfillment)?
- Hearing from students in their own words elevates their voices and humanizes their experiences.
- Students’ responses were primarily focused on physiological needs (food, housing) over other needs (safety, belonging, esteem, fulfillment). The second most frequently used of all words in 2024, 2023, and 2022 was literally “food.” Housing was also top-of-mind; the word “housing” was the 4th and 5th most frequently used word in 2024 and 2023, respectively.
- Multimethod content analysis of students’ responses from all three years found that students have the following needs:
  - Access to inclusive and nutritious food (e.g., more options and resources, free or cheaper food, an on-campus food pantry, SNAP EBT (food stamps) to be accepted on campus)
  - Affordable housing and education (e.g., lower on-campus housing costs, lower tuition, increased financial aid, more student employment opportunities with raised student wages)
  - Improved safety on- and off-campus (e.g., improved campus lighting, better security, extended Duck Rides hours)
  - Mental health and wellbeing support (e.g., increased therapist availability, no limit on number of therapy sessions, flexible attendance policies for mental health and illness, better education of faculty around mental health)
  - Community and inclusion (e.g., support for marginalized and neurodiverse students, improved DEI initiatives, more opportunities for social and community engagement)
  - Academic support and resources (e.g., advertisement of available supports and services, a one-stop support hub, faculty accountability)
Validity (see also Supplemental Method)
- The USDA food security measure has been extensively tested and in use for more than a quarter century.
- Independent assessment shows that the USDA measurement tends to underestimate, not overestimate, the prevalence of food insecurity (Gregory, 2020).
- Recent results from the National College Health Assessment (spring 2024) and the Hope Center Student Basic Needs Survey (winter 2023) at the UO suggest that our results do not overestimate food insecurity. The National College Health Assessment uses an adapted, shorter, and arguably less reliable version of the USDA module. Hope uses the same, 10-item form of the USDA module we use, though with a different recall time frame.
  - The National College Health Assessment (spring 2024) estimated that 45.1% (n = 223) of first-through-fourth-year U.S. undergraduates at UO were food insecure, compared to our estimate of 39.5(n = 487).
  - The Hope Center Student Basic Needs Survey estimated that 37.8% (n = 556) of first-through-fourth-year U.S. undergraduates at UO were food insecure, which is only slightly lower than our estimate in 2023 of 39.7% (n = 316). The estimates are almost exactly the same if 2 percentage points are deducted from our estimate for oversampling 2SLGBTQIA3+ students.
  - Additionally, Hope’s estimate of food insecurity among 2SLGBTQIA3+ students (43.4%), who are one of the more vulnerable groups according to our results, is also in line with ours (44.0%).
- The food insecurity rates at Oregon State University (Edwards et al., 2022), Portland State University (Townley et al., 2020), and Western Oregon University (Patton-Lopez et al., 2014), as well as national estimates from literature reviews and large national survey, also provide supporting evidence that the our results do not overestimate food insecurity (see What is Food Security section above).
- Observed higher rates of food insecurity among college students with lower socioeconomic status backgrounds, who identify as 2SLGBTQIA3+, are Latine/a/o/x, are from marginalized racial groups (e.g., Black, Indigenous), or are transfer students all confirm associations documented by previous research, as do observed associations between food security and stress, sadness, and life satisfaction (see What is Food Security section above).
- In our data, the patterns of relationships between food security and demographic/wellbeing variables are largely consistent across years, meaning that measurements of food security is reliable.
- It is reasonable to expect increased food insecurity (2024: 39.5, 2023: 39.7%; 2022: 34.0%; 2021: 21.8%) in the absence of greater action to mitigate it, given that food costs have risen dramatically. Nationally, food costs increased 3.5% in 2020, 3.9% in 2021, a whopping 9.9% in 2022, 5.8% in 2023, and they are expected to rise 2.3% (+/-.5%) across 2024 (USDA).
- Although we do welcome it, doubts about the validity of our results are unlikely based on earnest methodological critique. Rather, such doubts are more likely rooted in the fact that the results are rather unflattering. Even discounting the 2023 food insecurity rate because of likely oversampling of 2SLGBTQIA3+ students, the rate is still quite high. It is highly unlikely that any reasonable amount of discounting will make the rate low enough to justify ignoring it or not acting on it.

Recommendations

Undergraduates at the UO experience food insecurity at a much higher rate than the national household average, and that appears to be holding steady. Certain populations of students are even more food insecure. Rather than brushing these results aside, we should do more to lower the rate of food insecurity and help our student population.
Working against efforts to lower the rate of food insecurity, food costs have risen dramatically. Nationally, food costs increased 3.5% in 2020, 3.9% in 2021, a whopping 9.9% in 2022, 5.8% in 2023, and they are expected to rise 2.3% (+/-.5%) across 2024 (USDA). Contextually, the 10-year moving average of food price inflation from 2014 to 2023 was only 3.8% in the metropolitan area of the continental U.S. with the highest inflation: Tampa, FL (USDA). That is to say, current circumstances are cause for concern. With stagnant wage growth for low wage jobs, like those undergraduates tend to have, it is and will continue to be difficult to make ends meet.
As food insecurity is consistently associated with lower wellbeing (more stressed, sadder, less satisfied with life), food security should be considered as an important antecedent condition of student wellbeing and success. If students do not have reliable access to healthy, nutritious food, they are unlikely to be well or to succeed.
It is critical that food security programming and services be directed toward students with lower socioeconomic status backgrounds.
- Programming and services directed toward 2SLGBTQIA3+ students, Latine/a/o/x students, and Black, Native American/Alaska Native, and Native Hawaiian/Pacific Islander students may be effective in their own right. However, such an approach alone will not address the whole problem, given that it appears to be more of a broad socioeconomic issue than it is a specifically gender identity/sexuality one or a racial/ethnic one. With that said, because gender identity/sexuality tends to account for significant variance in food security above and beyond socioeconomic status indicators, it may be beneficial to target programming and services at 2SLGBTQIA3+ students in addition to students with lower socioeconomic status backgrounds.
- In open responses regarding basic needs support, 2SLGBTQIA3+ students raised a particular problem, namely that the food pantry is operated off-campus by a church. Given the national climate of discrimination against and dehumanization of 2SLGBTQIA3+ people, often by the religious right, it is not surprising that 2SLGBTQIA3+ students, who are one of the more vulnerable populations in terms of food security, may want to avoid religion-based resources. Perhaps establishment of a secular food pantry might be pursued.
Studying food security addresses one of many needs people have. The Student Wellbeing and Success Initiative helps us on this front, heavily emphasizes another need (belonging), and might be leveraged to at least partly address two others (self-esteem and fulfillment/self-actualization). However, it cannot measure everything. To more holistically understand and support students’ needs, the Student Wellbeing and Success Initiative itself needs more support and there needs to be more support for developing the nexus of knowledge through emphasis on studying other need domains.
- For example, in 2023, the Hope Center Student Basic Needs Survey estimated housing insecurity at 34.9% and homelessness at 11.1% among first-through-fourth-year U.S. undergraduates at the UO. Both housing insecurity and homelessness both slightly higher among 2SLGBTQIA3+ students (Housing Insecure: 40.2%; Experiencing Homelessness: 15.3%).

Overall Estimates

Food security was measured using the 10-item USDA Food Security Module. See Food Security Measurement for more detail.

2024

Note: n = 1,233. Error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 17,282 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2024. Numerically, the confidence interval the estimates was +/- 2.6 percentage points.

2023

Note: n = 795. It may be reasonable to decrease the estimated food insecurity (increase food security) rate by 2 percentage points, given likely oversampling of 2SLGBTQIA3+ students. Error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 16,778 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2023. Numerically, the confidence interval the estimates was +/- 3.3 percentage points.

2022

Note: n = 1,175. Error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 15,754 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2022. Numerically, the confidence interval the estimates was +/- 2.6 percentage points.

2021

Note: n = 790. Error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 12,563 admitted, active status, first-through-third-year, U.S. undergraduates during spring 2021. Numerically, the confidence interval the estimates was +/- 2.8 percentage points. The 2021 sample did not include fourth-year students, and students of marginalized racial groups (Black, Native American/Alaskan, Native Hawaiian/Pacific Islander) were undersampled. As such, the overall rate in 2021 may have been an underestimate.

By Demographics

Zero-Order Relationships

Food security was measured using the 10-item USDA Food Security Module. See Food Security Measurement for more detail.

2024

Disability/Neurodivergence

Students who have disabilities or are neurodivergent were significantly less likely to be food secure than were those who do not have disabilities or are not neurodivergent (b = -0.63, z(1,158) = -5.13, p < .001.

Note: n = 1,160 (n_2SLGBTQIA3+ = 612, n_{Not 2SLGBTQIA3+} = 548). Regressions were computed using the glm function with disability/neurodivergence dummy coded as Not Disabled/Neurodivergent = 0, Disabled/Neurodivergent = 1.

Gender Identity/Sexuality

2SLGBTQIA3+ students were significantly less likely to be food secure than were those who are not 2SLGBTQIA3+ (b = -0.39, z(1,177) = -3.26, p = .001).

Note: n = 1,179 (n_2SLGBTQIA3+ = 591, n_{Not 2SLGBTQIA3+} = 588). Regressions were computed using the glm function with gender identity/sexuality dummy coded as Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1.

Race/Ethnicity

Asian students were significantly more likely to be food secure than was the average student (b = 0.48, z(1,214) = 2.29, p = .022), and Latine/a/o/x students were significantly less likely to be food secure than was the average student (b = -0.47, z(1,214) = -3.52, p < .001. Black/Indigenous (b = -0.33, z(1,214) = -0.98, p = .326) and Multiracial/ethnic (b = -0.22, z(1,214) = -1.26, p = .207) students did not differ significantly on food security compared to the average student. However, small group size may have obscured a real difference. The odds of Black/Indigenous students being food insecure were 40% greater than those of the average student (OR = 1.40).

Note: n = 1,219 (n_Asian = 102, n_Latine/a/o/x 196, n_{Black/Indigenous} = 36 [69% Black], n_{Multiracial/ethnic} = 124, n_White = 761). Regressions were computed using the glm function with race/ethnicity being weighted effects coded (J. Cohen et al., 2003) with White as the base category, which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic were compared to “the average student,” weighted by group size.

Socioeconomic Status

Students of lower socioeconomic status were significantly less likely to be food secure than were those of higher socioeconomic status (b = -0.47, z(960) = -7.39, p < .001).

Note: n_{Food Security} = 962 (n_{Lower SES} = 191, n_Upper
SES = 771. Regressions were computed using the glm function with socioeconomic status retaining its full ordinal scale. That is, socioeconomic status is a composite variable composed of four indicators of lower socioeconomic status. More indicators corresponds to lower status, meaning that students with four indicators are the most marginalized and students with zero indicators are the most privileged in terms of socioeconomic status. For depicting results here, the scale was condensed to two categories: Lower SES (2-4 Indicators) and Upper SES (0-1 Indicators). See Supplemental Method for details.

Transfer Status

Transfer students were significantly less likely to be food secure than were first-time students (b = -0.35, z(1,225) = -1.98, p = .048).

Note: n = 1,227 (n_Transfer = 142, n_First-Time = 1,085). Regressions were computed using the glm function with transfer status dummy coded as First-Time = 0, Transfer = 1.

2023

Disability/Neurodivergence

Students who have disabilities or are neurodivergent were significantly less likely to be food secure than were those who do not have disabilities or are not neurodivergent (b = -0.79, z(732) = -4.80, p < .001.

Note: n = 734 (n_2SLGBTQIA3+ = 460, n_{Not 2SLGBTQIA3+} = 274). Regressions were computed using the glm function with disability/neurodivergence dummy coded as Not Disabled/Neurodivergent = 0, Disabled/Neurodivergent = 1.

Gender Identity/Sexuality

2SLGBTQIA3+ students were significantly less likely to be food secure than were those who are not 2SLGBTQIA3+ (b = -0.31, z(763) = -2.05, p = .041).

Note: n = 765 (n_2SLGBTQIA3+ = 455, n_{Not 2SLGBTQIA3+} = 310). Regressions were computed using the glm function with gender identity/sexuality dummy coded as Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1.

Race/Ethnicity

Latine/a/o/x students were marginally less likely to be food secure than was the average student (b = -0.32, z(779) = -1.82, p = .068). Asian (b = 0.37, z(779) = 1.05, p = .294), Multiracial/ethnic (b = -0.38, z(779) = -1.58, p = .114), and Black/Indigenous (b = -0.56, z(779) = -1.45, p = .147) students did not differ significantly on food security compared to the average student. However, small group size may have obscured some real differences. The odds of Asian students being food secure were 45% greater than those of the average student (OR = 1.45), whereas the odds of Multiracial/ethnic students being food insecure were 46% greater than those of the average student (OR = 1.46), and the odds of Black/Indigenous (Black, Native American/Alaska Native, and Native Hawaiian/Pacific Islander) students being food insecure were 75% greater than those of the average student (OR = 1.75). Conversely and illustratively, Latine/a/o/x students were the largest group and their effect size was bordering on being substantial (OR = 1.38), like their p-value was bordering on being significant (p = = .068).

Note: n = 784 (n_Asian = 35, n_Latine/a/o/x 113, n_{Black/Indigenous} = 26 [77% Black], n_{Multiracial/ethnic} = 65, n_White = 545). Regressions were computed using the glm function with race/ethnicity being weighted effects coded (J. Cohen et al., 2003) with White as the base category, which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic were compared to “the average student,” weighted by group size.

Socioeconomic Status

Students of lower socioeconomic status were significantly less likely to be food secure than were those of higher socioeconomic status (b = -0.45, z(584) = -5.33, p < .001).

Note: n_{Food Security} = 586 (n_{Lower SES} = 114, n_Upper
SES = 472. Regressions were computed using the glm function with socioeconomic status retaining its full ordinal scale. That is, socioeconomic status is a composite variable composed of four indicators of lower socioeconomic status. More indicators corresponds to lower status, meaning that students with four indicators are the most marginalized and students with zero indicators are the most privileged in terms of socioeconomic status. For depicting results here, the scale was condensed to two categories: Lower SES (2-4 Indicators) and Upper SES (0-1 Indicators). See Supplemental Method for details.

Transfer Status

Transfer students were significantly less likely to be food secure than were first-time students (b = -0.50, z(792) = -2.58, p = .010).

Note: n = 794 (n_Transfer = 130, n_First-Time = 664). Regressions were computed using the glm function with transfer status dummy coded as First-Time = 0, Transfer = 1.

2022

Disability/Neurodivergence

[Insert blurb here explaining why this isn’t reported for this year?]

Gender Identity/Sexuality

2SLGBTQIA3+ students were significantly less likely to be food secure than were those who are not 2SLGBTQIA3+ (b = -0.42, z(1,142) = -3.34, p < .001).

Note: n = 1,144 (n_2SLGBTQIA3+ = 504, n_{Not 2SLGBTQIA3+} = 640). Regressions were computed using the glm function with gender identity/sexuality dummy coded as Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1.

Race/Ethnicity

Asian students were significantly more likely to be food secure than was the average student (b = 0.50, z(1,146) = 2.07, p = .038). Latine/a/o/x (b = -0.39, z(1,146) = -2.59, p = .010) and Black/Indigenous (b = -0.76, z(1,146) = -2.36, p = .018) students were significantly less likely to be food secure than was the average student. Multiracial/ethnic students did not differ significantly on food security compared to the average student (b = -0.14, z(1,146) = -0.72, p = .472).

Note: n = 1,151 (n_Asian = 87, n_Latine/a/o/x 156, n_{Black/Indigenous} = 38 [76% Black], n_{Multiracial/ethnic} = 104, n_White = 766). Regressions were computed using the glm function with race/ethnicity being weighted effects coded (J. Cohen et al., 2003) with White as the base category, which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic were compared to “the average student,” weighted by group size.

Socioeconomic Status

Students of lower socioeconomic status were significantly less likely to be food secure than were those of higher socioeconomic status (b = -0.39, z(810) = -5.37, p < .001).

Note: n_{Food Security} = 812 (n_{Lower SES} = 161, n_Upper
SES = 651. Regressions were computed using the glm function with socioeconomic status retaining its full ordinal scale. That is, socioeconomic status is a composite variable composed of four indicators of lower socioeconomic status. More indicators corresponds to lower status, meaning that students with four indicators are the most marginalized and students with zero indicators are the most privileged in terms of socioeconomic status. For depicting results here, the scale was condensed to two categories: Lower SES (2-4 Indicators) and Upper SES (0-1 Indicators). See Supplemental Method for details.

Transfer Status

Transfer students were significantly less likely to be food secure than were first-time students (b = -0.50, z(1,171) = -2.90, p = .004).

Note: n = 1,173 (n_Transfer = 163, n_First-Time = 1,010). Regressions were computed using the glm function with transfer status dummy coded as First-Time = 0, Transfer = 1.

2021

Disability/Neurodivergence

[Insert blurb here explaining why this isn’t reported for this year?]

Gender Identity/Sexuality

2SLGBTQIA3+ students were significantly less likely to be food secure than were those who are not 2SLGBTQIA3+ (b = -0.37, z(778) = -2.09, p = .036).

Note: n = 780 (n_2SLGBTQIA3+ = 303, n_{Not 2SLGBTQIA3+} = 477). Regressions were computed using the glm function with gender identity/sexuality dummy coded as Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1.

Race/Ethnicity

Asian (b = (b = 0.10, z(758) = 0.31, p = .754), Latine/a/o/x (b = -0.17, z(758) = -0.78, p = .435), Multiracial/ethnic (b = -0.12, z(758) = -0.43, p = .668), and Black/Indigenous (Black, Native American/Alaska Native, and Native Hawaiian/Pacific Islander) (b = -0.95, z(758) = -1.63, p = .103) students did not differ significantly on food security compared to the average student. However, small group size may have obscured a real difference. The odds of Black/Indigenous students being food insecure were more than twice those of the average student (OR = 2.58). Other odds ratios were very small/negligible.

Note: n = 763 (n_Asian = 55, n_Latine/a/o/x 97, n_{Black/Indigenous} = 12 [83% Black], n_{Multiracial/ethnic} = 63, n_White = 536). Regressions were computed using the glm function with race/ethnicity being weighted effects coded (J. Cohen et al., 2003) with White as the base category, which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic were compared to “the average student,” weighted by group size.

Socioeconomic Status

Students of lower socioeconomic status were significantly less likely to be food secure than were those of higher socioeconomic status (b = -0.24, z(663) = -3.10, p = .002).

Note: n_{Food Security} = 665 (n_{Lower SES} = 149, n_Upper
SES = 516. Regressions were computed using the glm function with socioeconomic status retaining its full ordinal scale. That is, socioeconomic status is a composite variable composed of four indicators of lower socioeconomic status. More indicators corresponds to lower status, meaning that students with four indicators are the most marginalized and students with zero indicators are the most privileged in terms of socioeconomic status. For depicting results here, the scale was condensed to two categories: Lower SES (2-4 Indicators) and Upper SES (0-1 Indicators). See Supplemental Method for details.

Transfer Status

Transfer students did not differ significantly on food security compared to first-time students (b = -0.34, z(753) = -1.45, p = .148).

Note: n = 755 (n_Transfer = 114, n_First-Time = 641). Regressions were computed using the glm function with transfer status dummy coded as First-Time = 0, Transfer = 1.

Accounting for Unique Variation

Food security was measured using the 10-item USDA Food Security Module.

2024

When the demographic characteristics that significantly or at least substantially predicted food security (disability/neurodivergence, gender identity/sexuality, race/ethnicity, socioeconomic status, transfer status) were all entered into one model, disability/neurodivergence (b = -0.53, z(909) = -3.58, p < .001), race/ethnicity, specifically the comparison of Asian students to the average student (b = 0.74, z(909) = 2.71, p = .007), and socioeconomic status (b = -0.45, z(909) = -6.57, p < .001), remained significant. None of the nonsignificant predictors were of substantial size.

Note: Predictor variables were disability/neurodivergence (dummy coded – Not Disabled/Neurodivergent = 0, Disabled/Neurodivergent = 1) gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1).

2023

When the demographic characteristics that significantly or at least substantially predicted food security (disability/neurodivergence, gender identity/sexuality, race/ethnicity, socioeconomic status, transfer status) were all entered into one model, disability/neurodivergence (b = -0.86, z(538) = -4.03, p < .001) and socioeconomic status (b = -0.44, z(538) = -4.74, p < .001), remained significant.

The nonsignificant predictors of food security that were nonetheless of substantial size were the comparisons of Asian and Black/Indigenous students to the average student. The odds of Asian students being food secure were 71% greater than the average student (OR = 1.71), and the odds of Black/Indigenous students being food insecure were 71% greater than those of the average student (OR = 1.71).

Note: Predictor variables were disability/neurodivergence (dummy coded – Not Disabled/Neurodivergent = 0, Disabled/Neurodivergent = 1) gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1).

2022

When the demographic characteristics that significantly or at least substantially predicted food security (gender identity/sexuality, race/ethnicity, socioeconomic status, transfer status) were all entered into one model, gender identity/sexuality (b = -0.41, z(781) = -2.64, p = .008) and socioeconomic status (b = -0.36, z(781) = -4.70, p < .001), remained significant.

Additionally, Black/Indigenous students were marginally less likely to be food secure compared to the average student, when controlling for gender identity/sexuality, socioeconomic status, and transfer status (b = -0.86, z(781) = -1.82, p = .069). The odds of Black/Indigenous students being food insecure were more than twice those of the average student (OR = 2.36).

The nonsignificant predictors that were nonetheless of substantial size were the comparison of Asian students (b = 0.44, z(781) = 1.57, p = .117) and transfer students (b = -0.42, z(781) = -1.47, p = .142) to the average student. The odds of Asian students being food secure were 55% greater than those of the average student (OR = 1.55). The odds of transfer students being food insecure were 52% greater than those of the average student (OR = 1.52).

Note: Predictor variables were gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1). Disability/neurodivergence was not collected during this timeframe.

2021

When the demographic characteristics that significantly or at least substantially predicted food security (gender identity/sexuality, race/ethnicity, socioeconomic status, transfer status) were all entered into one model, only socioeconomic status remained significant (b = -0.20, z(627) = -2.33, p = .020). Additionally, gender identity/sexuality was marginally significant (b = -0.37, z(627) = -1.88, p = .060).

There were two nonsignificant predictors that were nonetheless of substantial size: the comparison of Multiracial/ethnic (b = -0.31, z(627) = -1.05, p = .292) and Black/Indigenous (b = 0.35, z(627) = 0.31, p = .758) students to the average student (b = -0.31, z(627) = -1.05, p = .292). The odds of Multiracial/ethnic students being food insecure were 37% greater than the average student (OR = 1.37). The coefficient for Black/Indigenous reversed direction in the multiple-predictor model, meaning that when the other demographic variables were accounted for, Black/Indigenous students were more likely to be food secure compared to the average student. Specifically, the odds of Black/Indigenous students being food secure were 41% greater than those of the average student (OR = 1.41). These results suggest the model may be untrustworthy.

Note: Predictor variables were gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Black/Indigenous, Latine/a/o/x, and Multiracial/ethnic are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1).

Wellbeing Consequences

Food security was measured using the 10-item USDA Food Security Module. See Food Security Measurement for more detail.

2024

Stress

Students who were food secure were significantly less stressed (b = -8.96, t (1,210) = -9.12, p < .001). This relationship remained significant (b = -7.50, t (909) = -6.48, p < .001) after controlling for disability/neurodivergence, gender identity/sexuality, socioeconomic status, and transfer status, which were potential confounding variables given their relationships with both stress and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and stress. Control covariates were disability/neurodivergence (dummy coded – No Disability/Neurodivergence = 0, Has Disability/Neurodivergence = 1), gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Sadness

Students who were food secure were significantly less sad (b = -9.72, t (1,202) = -6.76, p < .001. This relationship remained significant (b = -6.85, t (888) = -4.07, p < .001) after controlling for disability/neurodivergence, gender identity/sexuality, race/ethnicity, socioeconomic status, and transfer status, which were potential confounding variables given their relationships with both sadness and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and sadness. Control covariates were disability/neurodivergence (dummy coded – No Disability/Neurodivergence = 0, Has Disability/Neurodivergence = 1), gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Latine/a/o/x, Multiracial/ethnic, and Black/Indigenous are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Life Satisfaction

Students who were food secure were significantly more satisfied with life (b = 11.18, t (1,218) = 8.54,< .001). This relationship remained significant (b = 9.24, t (900) = 6.03, p < .001) after controlling for disability/neurodivergence, gender identity/sexuality, race/ethnicity, socioeconomic status, and transfer status, which were potential confounding variables given their relationships with both life satisfaction and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and life satisfaction. Control covariates were disability/neurodivergence (dummy coded – No Disability/Neurodivergence = 0, Has Disability/Neurodivergence = 1), gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Latine/a/o/x, Multiracial/ethnic, and Black/Indigenous are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

2023

Stress

Students who were food secure were significantly less stressed (b = -9.23, t (760) = -6.91, p < .001). This relationship remained significant (b = -6.97, t (524) = -4.18, p < .001) after controlling for disability/neurodivergence, gender identity/sexuality, race/ethnicity, and socioeconomic status which were potential confounding variables given their relationships with both stress and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and stress. Control covariates were disability/neurodivergence (dummy coded – No Disability/Neurodivergence = 0, Has Disability/Neurodivergence = 1), gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Latine/a/o/x, Multiracial/ethnic, and Black/Indigenous are compared to “the average student,” weighted by group size), and socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Sadness

Students who were food secure were significantly less sad (b = -10.41, t (773) = -5.38, p < .001. This relationship remained significant (b = -5.81, t (529) = -2.43, p = .015) after controlling for disability/neurodivergence, gender identity/sexuality, race/ethnicity, and socioeconomic status, which were potential confounding variables given their relationships with both sadness and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and sadness. Control covariates were disability/neurodivergence (dummy coded – No Disability/Neurodivergence = 0, Has Disability/Neurodivergence = 1), gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Latine/a/o/x, Multiracial/ethnic, and Black/Indigenous are compared to “the average student,” weighted by group size), and socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Life Satisfaction

Students who were food secure were significantly more satisfied with life (b = 9.70, t (772) = 5.76,< .001). This relationship remained significant (b = 6.88, t (531) = 3.35, p < .001) after controlling for disability/neurodivergence, gender identity/sexuality, race/ethnicity, socioeconomic status, and transfer status, which were potential confounding variables given their relationships with both life satisfaction and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and life satisfaction. Control covariates were disability/neurodivergence (dummy coded – No Disability/Neurodivergence = 0, Has Disability/Neurodivergence = 1), gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Latine/a/o/x, Multiracial/ethnic, and Black/Indigenous are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

2022

Stress

Students who were food secure were significantly less stressed (b = -10.79, t (1,151) = -9.10, p < .001). This relationship remained significant (b = -8.83, t (788) = -5.95, p < .001) after controlling for gender identity/sexuality and socioeconomic status which were potential confounding variables given their relationships with both stress and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and stress. Control covariates were gender identity/sexuality and socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Sadness

Students who were food secure were significantly less sad (b = -11.36, t (1,146) = -6.86, p < .001. This relationship remained significant (b = -9.17, t (783) = -4.49, p < .001) after controlling for gender identity/sexuality and socioeconomic status, which were potential confounding variables given their relationships with both sadness and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and sadness. Control covariates were gender identity/sexuality and socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Life Satisfaction

Students who were food secure were significantly more satisfied with life (b = 13.05, t (1,148) = 8.91,< .001). This relationship remained significant (b = 9.13, t (787) = 5.09, p < .001) after controlling for gender identity/sexuality, socioeconomic status, and transfer status, which were potential confounding variables given their relationships with both life satisfaction and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and life satisfaction. Control covariates were gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

2021

Stress

Students who were food secure were significantly less stressed (b = -9.23, t (782) = -5.90, p < .001). This relationship remained significant (b = -8.95, t (743) = -5.68, p < .001) after controlling for gender identity/sexuality, which was a potential confounding variable given its relationships with both stress and food security, and race/ethnicity, which was used as precision control covariate given its relationships with stress but not food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and stress. Control covariates were gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1) and race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Latine/a/o/x, Multiracial/ethnic, and Black/Indigenous are compared to “the average student,” weighted by group size). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Sadness

Students who were food secure were significantly less sad (b = -13.58, t (786) = -5.84, p < .001. This relationship remained significant (b = -13.21, t (776) = -5.69, p < .001) after controlling for gender identity/sexuality, which was a potential confounding variable given its relationship with both sadness and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and sadness. The control covariate was gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Life Satisfaction

Students who were food secure were significantly more satisfied with life (b = 10.55, t (783) = 5.37,< .001). This relationship remained significant (b = 8.08, t (623) = 3.75, p < .001) after controlling for gender identity/sexuality, race/ethnicity, socioeconomic status, and transfer status, which were potential confounding variables given their relationships with both life satisfaction and food security.

Note: Means depicted are from controlled models. Error bars represent 95% confidence intervals around means. Control covariates were chosen based on the sizes of zero-order relationships between demographics and food security and between demographics and life satisfaction. Control covariates were gender identity/sexuality (dummy coded – Not 2SLGBTQIA3+ = 0, 2SLGBTQIA3+ = 1), race/ethnicity (weighted effects coded J. Cohen et al., 2003 with White as the base category, which means that which means that Asian, Latine/a/o/x, Multiracial/ethnic, and Black/Indigenous are compared to “the average student,” weighted by group size), socioeconomic status (ordinal composite scale: 0 to 4 indicators of lower socioeconomic status; see Supplemental Method for details), and transfer status (dummy coded – First-Time = 0, Transfer = 1). Regressions were computed using the lm function and model predicted means and 95% confidence intervals were computed using the emmeans package.

Student Voices – Basic Needs Support

Results

2024

Hierarchy of Needs

In spring 2022, we introduced a general open-ended question: How can the UO better support your basic needs (food, housing, safety, belonging, self-esteem, fulfillment)? In response, students had the most to say about physiological needs. A total of 165 students (28.8%) explicitly mentioned the word “food” 190 times. A total of 69 students (12.1%) explicitly mentioned the word “housing” 85 times. These frequencies underestimate the extent to which students are writing about housing or food issues, given that many other words can be and are used to describe housing of food issues. Still, “food” is the second most frequently used word and “housing” is the fourth most frequently used word, overall, which puts both domains of issues top-of-mind.

The problem is that many students struggle to afford food or housing. A large portion (50.3%) of students mentioned financially related terms (499 instances), with 13.2% of them mentioning the more specific terms “pay,” “paid,” or “wage.”

Students mentioned all other domains of Maslow’s hierarchy but much less frequently. A total of 15 students (2.6%) mentioned some version of the word “safe” 18 times. A total of 1 students (0.2%) mentioned some version of the word “belong” 1 times. A total of 3 students (0.5%) mentioned the word “esteem” 3 times. A total of 1 students (0.2%) mentioned some version of the word “fulfill” 1 times.

Note: The “Physiological” category combines mentions of the terms “food” and “housing.” Percentages were computed dividing the number of students who mentioned a given term (or terms in the case of “Physiological”) by the number of responses used in analysis and multiplying by 100.

Content Analysis

Themes

Overview

A group of students (n = 59) felt that their basic needs were being supported well by UO, acknowledged their privilege that their needs were otherwise satisfied, or felt that their own need satisfaction was their personal responsibility and not the university’s. However, a staggering number of students were struggling with meeting their needs or saw room for improvement in the university’s support systems (n = 572, 45.6% of the sample). Students were candid and shared how they thought the UO could improve, which we summarized in six categories of themes, described below. Themes can be thought of as both conclusions we drew from what students told us and recommendations to staff and faculty for action.

Access to Inclusive and Nutritious Food

This category focuses on ensuring all students have access to inclusive and nutritious food, which is crucial for their overall wellbeing and academic success. This means providing affordable and healthier food options that are culturally inclusive and account for dietary needs (e.g., dairy-free, gluten-free, vegan, vegetarian) in campus dining halls and venues, establishing an on-campus food pantry for students, increasing produce drops, and implementing discounted dining plans for student employees and those with dietary restrictions. Additionally, accepting SNAP benefits (i.e., food stamps) on campus and offering more meal points can significantly alleviate food insecurity among students. Educational initiatives like cooking classes and nutrition workshops can empower students to make healthier choices and manage their food budgets effectively.

Housing Affordability

This category addresses the financial burden of housing for students in Eugene, where rising rents make finding affordable accommodations difficult. Lowering on-campus housing costs and offering more scholarships can help. UO can work with local authorities and landlords to develop affordable options and secure reduced rent agreements. A dedicated housing office can assist with leases, tenant rights, and housing searches would be ideal. UO can implement stricter security measures, such as increased lighting, regular patrols, and secure access systems by campus housing. Addressing hygiene issues in on-campus housing should be a top priority, with students experiencing illnesses due to food poisoning. Guaranteeing that housing is safe, accessible, and inclusive for all students, including those with disabilities, will create a supportive and secure living environment conducive to academic and personal success.

College Affordability

This category encompasses all the issues related to the financial burdens faced by students, including high tuition and inadequate student employment opportunities. It includes scholarships, grants, financial aid programs, basic needs support, affordable tuition initiatives, and support services that help students manage and reduce the cost of college. The focus is on making higher education accessible to all students, regardless of their economic background, by providing financial resources, counseling, and tools to help them navigate college expenses. Efforts within this category aim to reduce student debt, provide transparent information about costs, and promote financial literacy to ensure students can afford and complete their education without undue financial burden. Addressing the high cost of tuition and other related expenses is essential to make college more affordable for all students to include transportation costs. Lowering tuition rates for both in-state and out-of-state students can significantly reduce financial stress. Increasing student wages and providing more financial aid and scholarships will help cover the various costs associated with college. Offering grants and financial support for expenses beyond tuition, such as books and supplies, can further alleviate the financial burden on students. Utilizing donations from alumni and boosters to support non-athlete students can provide additional financial resources and opportunities for those in need.

Safety

This category focuses on ensuring the physical safety and security of students on and off campus. It includes concerns and suggestions related to campus lighting, security measures, transportation safety, and overall campus environment. The goal is to create a safe and secure atmosphere where students feel protected and supported, both during the day and at night. Students indicate the importance of adequate security personnel, effective safety protocols, and accessible transportation options to enhance students’ sense of safety and wellbeing. Expanding Duck Rides to operate during all hours of darkness can enhance safety. Protecting the safety of students on and around campus can foster a secure and supportive educational environment. Additionally, enhancing emergency response protocols and providing clear communication about safety resources can address concerns about violence, theft, and other dangers. Creating safe housing options near campus for those who cannot afford to live on campus, particularly for marginalized students, and promoting a culture of respect and inclusion assure that all students feel safe and supported in their daily lives.

Mental Health and Wellbeing

This category addresses the mental health challenges students face, including the availability and accessibility of counseling services, managing academic stress, and ensuring physical health. It emphasizes the need for comprehensive mental health support and effective safety measures. Normalizing mental health issues and reducing their stigma are essential for fostering a supportive environment. Flexible attendance policies can accommodate students dealing with mental health struggles or illness. Expanding mental health services and support systems is a component of improving student wellbeing. Removing the cap on therapy sessions and providing information about local therapists and group therapy options will ensure continuous access to mental health care. Increasing the number of counselors and improving the accessibility of mental health resources can meet the growing demand for these services. Establishing regular mental health check-ins, offering emergency mental health support, and providing mental health days can help students manage stress and maintain their mental health. Educating professors on how to support students with mental health issues will create a more understanding and compassionate academic environment.

Community and Inclusion

This category addresses the social and inclusion-related challenges students face, such as difficulties in social integration, lack of support for marginalized and neurodiverse students, and the need for a more inclusive campus environment. It emphasizes the importance of creating a welcoming and supportive community for all students, that everyone, regardless of their background, has access to the resources and support needed to succeed. An approach to inclusion encompasses services, programs, and initiatives aimed at academic support, mental health services, cultural and identity-based resources, and accommodations for students with disabilities. Supporting marginalized students and promoting equity can help create a sense of belonging. Providing tailored resources and programs for students with children, nontraditional students, commuter students and those with disabilities helps these students feel supported, included, and can help them connect with their peers.

Academic Support and Resources

This category focuses on the academic challenges students encounter, emphasizing the need for improved tutoring, advising, and course flexibility. A peer-to-peer model can enhance learning experiences, while offering flexible makeup work and assignment deadlines can accommodate diverse student needs. Enhancing the awareness and accessibility of academic support services is vital for creating a supportive learning environment. Fostering a culture of understanding and compassion among faculty, along with revising attendance policies to accommodate student needs, can create a more inclusive academic environment. Potential changes to the term system can provide students with the adaptability they need to thrive. Establishing a one-stop hub, whether in-person or online, for all student support services can streamline the process of seeking assistance. Providing additional tutoring and study resources, especially for challenging subjects, will help students overcome academic obstacles and achieve their educational goals.

Frequencies

Theme	Description	n	%
Access to Inclusive and Nutritious Food	This category focuses on ensuring all students have access to inclusive and nutritious food, which is crucial for their overall wellbeing and academic success. This means providing affordable and healthier food options that are culturally inclusive and account for dietary needs (e.g., dairy-free, gluten-free, vegan, vegetarian) in campus dining halls and venues, establishing an on-campus food pantry for students, increasing produce drops, and implementing discounted dining plans for student employees and those with dietary restrictions. Additionally, accepting SNAP benefits (i.e., food stamps) on campus and offering more meal points can significantly alleviate food insecurity among students. Educational initiatives like cooking classes and nutrition workshops can empower students to make healthier choices and manage their food budgets effectively.	248	43.4

Housing Affordability	This category addresses the financial burden of housing for students in Eugene, where rising rents make finding affordable accommodations difficult. Lowering on-campus housing costs and offering more scholarships can help, while partnering with local authorities and landlords can secure reduced rent agreements. Implementing stricter security measures, addressing hygiene issues, and ensuring housing is safe, accessible, and inclusive will create a supportive environment conducive to student success.	76	13.3

College Affordability	This category addresses the financial burdens faced by students, including high tuition and inadequate employment opportunities. It covers scholarships, grants, financial aid programs, and support services to help manage and reduce college costs. The goal is to make higher education accessible to all by providing financial resources, counseling, and tools to navigate expenses, ultimately reducing student debt. Efforts include lowering tuition rates, increasing student wages, and offering additional financial support for expenses like books and supplies.	209	36.5

Safety	This category focuses on ensuring the physical safety and security of students on and off campus, addressing concerns related to campus lighting, security measures, transportation safety, and overall campus environment. The goal is to create a secure atmosphere where students feel protected and supported at all times. Enhancing emergency response protocols, expanding Duck Rides to operate during all hours of darkness, and promoting a culture of respect and inclusion are essential for fostering a safe and supportive educational environment.	24	4.2

Mental Health and Wellbeing	This category addresses the mental health challenges students face, focusing on the availability and accessibility of counseling services, managing academic stress, and ensuring physical health. It emphasizes comprehensive mental health support, normalizing mental health issues, and reducing stigma. Flexible attendance policies and expanding mental health services are essential for student wellbeing. Increasing the number of counselors, removing caps on therapy sessions, and providing regular mental health check-ins and emergency support can help students manage stress and maintain their mental health.	50	8.7

Community and Inclusion	This category addresses social and inclusion-related challenges, such as difficulties in social integration and lack of support for marginalized and neurodiverse students. It emphasizes creating a welcoming and supportive community where all students have access to necessary resources. This includes academic support, mental health services, cultural and identity-based resources, and accommodations for students with disabilities. Tailored programs for students with children, nontraditional students, commuters, and those with disabilities help foster a sense of belonging and connection among peers.	70	12.2

Academic Support and Resources	This category addresses academic challenges, emphasizing the need for improved tutoring, advising, and course flexibility. A peer-to-peer model, flexible makeup work, and assignment deadlines can accommodate diverse student needs. Enhancing the accessibility and awareness of academic support services and fostering a compassionate faculty culture can create a more inclusive environment. Establishing a one-stop hub for all student support services can streamline assistance and help students overcome academic obstacles.	105	18.4

Note: Themes can be thought of as both conclusions we drew from what students told us and recommendations to staff and faculty for action. Responses were open-coded using Grounded Theory (Strauss & Corbin, 2006; Tracy, 2019), which is a framework used to understand people’s experiences by building on themes from qualitative data they provide. Quotes were silently edited for clarity.

Wordnet

There are some interesting threads in the outlying 2-to-3-term clusters of the network depiction of adjacent adjectives and nouns below, but the main story is in the large cluster with the “food” and “more” centroids. Students are concerned about food security and want more options, supports, and resources related to basic needs. These include an improved food pantry situation (on campus, expanded hours, better and more inclusive options), affordable housing, lower tuition, more scholarships, more affordable food options on campus for students who don’t have meal plans, subsidized meal plans, and free Duck Bucks for those who need them. The smaller clusters show that students want (a) increased and improved mental health services, (b) more frequent produce drops,(c) affordable, nutritious, and inclusive meals and expanded hours at the dining halls, and (d) increased funding for the textbook subsidy program.

Note: Wordnet = adjacent ($\le$ 1 word apart) adjective-noun bigrams visualized as a network. Line thickness represents frequency of cooccurrence. The criterion for inclusion of bigram in the wordnet was cooccurrence $\gt$ 3 (min = 4, max = 39). If a word seems positively valenced, it may have been negated (e.g., preceded by the word “not”). Text data were annotated using the udpipe package.

Wordcloud

Keywords extracted and displayed in the wordcloud below largely confirm the descriptions in the Hierarchy of Needs, Themes, and Wordnet sections. Students are struggling to make ends meet and want more tangible support of their physiological needs, particularly food and housing.

Note: Word size represents frequency of keyword occurrence. Hover over words to see frequencies. The criterion for inclusion of a keyword in the wordcloud was occurrence $\gt$ 3. If a word seems positively valenced, it may have been negated (e.g., preceded by the word “not”). Text data were annotated using the udpipe package. Keywords were extracted using the Rapid Keyword Extraction (RAKE) algorithm (Rose et al., 2010).

Quotes

“Commit to funding the basic needs program more, same with committing to fund an on-campus food pantry.”

“Give more funding and support to the basic needs program. They have amazing resources, but I fear they do not have enough to keep up with demand.”

“UO desperately needs an on-campus UO -supported food pantry. There is need. Student orgs should not have to petition for it. It is way overdue.”

“Increased capacity and funding for the Basic Needs program so that they can cover more funds for more people. Increase funds for the Textbook Subsidy program and the Ducks Feeding Ducks program. Create an on-campus food pantry that is open at least 8 hours a day, preferably 10-12 hours. The food pantry should have quality food that is not expired and is good for many different dietary preferences and needs, such as gluten-free, dairy-free, and vegetarian/vegan options.”

“The food pantry has been a lifesaver before, but it needs to be available more than once a week. Some students work Tuesday or only on-campus Monday. Make it more accessable to people of all schedules.”

“When I was at PCC (Portland Community College), I felt a sense that the administration and faculty were focused on supporting student needs, especially for low-income and disadvantaged students. At UO, it often feels like supporting students is not the first priority and comes second to prestige and financial success.”

“The student sustainability Produce Drop is everything to a lot of students. They must be supported more.”

“I wish the Produce Drop could happen at different times. It always seems to be when I am at work.”

“I would suggest that the Produce Drop happen multiple times a week. My Tuesday schedule, and many of my friends’ Tuesday schedules, are very busy and keep us from accessing this.”

“Let us use food stamps on campus.”

“More information and access for students to resources like SNAP Oregon EBT would be extremely helpful.”

“Assistance for students to get onto SNAP needs to be available and well publicized. For those who do not qualify, more could be done to advertise the student food pantry.”

“Implement programs that educate students on healthy eating habits, and provide access to nutritious food options on campus.”

“As a person with many dietary restrictions, I find it difficult to get food that is safe for me to eat, nourishing, and fits within my meal points. I understand that gluten-free and vegan food is normally more expensive in groceries stores, but if UO is trying to better accommodate students with health needs, they should make gluten-free, vegan, dairy-free foods the same prices as their normal counterparts. For many people, it is not a decision to have these dietary restrictions, so they should not be disadvantaged because of it. In the dorm markets especially, gluten-free food is often at least two points more expensive and has less in it. So if the UO wants to make all people feel included, especially those with health needs, they should look to see where their dining services are falling short.”

“Food with extremely high grocery prices. I find that often I am skipping meals. When I go home for my checkups, I experience nutrient deficiencies that I never have in my life.”

“The food available in cafes or shops on campus should not be charging such large prices. A sandwich is around $9-$10, and if I want a tea or something, I am almost at $20 for one day for one person.”

“Provide more nutritious and various dining options around campus, and expand funding for Ducks Feeding Ducks. It’s a genuine lifesaver.”

“Limiting Ducks Feeding Ducks to once a term makes it harder. When it was three times per term, it was a great source for when I did not have money and needed food. I could make a Chipotle bowl or Subway last the entire day for lunch and dinner.”

“Allow more than one Ducks Feeding Ducks per person per term. Food insecurity is real at this school.”

“I would love for the Duck Bucks 50-dollar credit to extend to all students, not just freshmen. This would make it easier to get food on campus when I have not eaten, without spending the money I have been saving for groceries.”

“Make it so students off campus have Duck Bucks. We get hungry on campus, too.”

“I think the point system is pretty messed up in terms of the money-to-point ratio.”

“I feel like the meal points are a scam and should be based of actual money and not points. It really confusing, and I wish it was different. OSU bases their meal plans on money, not points.”

“Offer meal vouchers for students in need who do not have a meal plan.”

“Offer a rolling Textbook Subsidy program.”

“Offering lower tuition and cheaper housing with more scholarships and on-campus jobs; it is really difficult to find on-campus work that pays enough to cover living expenses.”

“Pay student workers more, and pay them biweekly.”

“Increase the expected living cost so that more financial aid can be used. The amount that housing costs in this town, along with other bills, is much higher than what is estimated by financial aid.”

“Estimates for housing fees need to be adjusted to be higher. Off-campus housing is becoming the norm for students, and rent is increasing astronomically. I have complained multiple times to financial aid about their housing estimates being far too low resulting in financial aid packages that allegedly cover my needs for the year but frankly fall short, despite having three roommates and living in a modest apartment. Students need more housing assistance and help finding adequate appropriate housing.”

“Safety mainly is the most important thing, and feeling safe when out and about at night.”

“Add some more street lights. Some places on campus get pretty dark at night.

“Add brighter lights next to the art museum and Knight Library.”

“For safety, the campus needs to be more well lit along 13th St.”

“Have more resources for safety on campus at night.”

“Traffic around campus feels really unsafe sometimes due to plants or trees that block the view of drivers, especially west of campus on Hilyard/Patterson. There have been multiple times when I was nearly hit while walking home from class because there is a lack of sidewalks on streets that lots of students use to walk to class, like 15th and 16th. Cars drive so fast or turn without checking for pedestrians. I don’t know if there a way for the school to advocate for something to help with that or not, but it would be smart to look into it.”

“Try our best to use resocures to clean up the surrounding commuity so it can be safer in general for all students.”

“Make it easier to get therapy appointments. It is so hard to get in touch with the health center. Even when you make an online appointment, they always say I did it wrong or someone was unavailable.”

“Make campus therapy appointments that aren’t limited.”

“I wish there were more than seven short-term therapy sessions per school year.”

“More availability at the counseling center. Whenever I have called to get help and it’s the middle of the term, they say they can’t really help me because everyone is already booked. More mental health services in general would be great, but especially those that do not require students to be on top of it starting the first week of the term.”

“I think that making sure all the students know about wellbeing coaching and other free mental health services on campus would be helpful, as well as housing options that are not as expensive because that is super hard to find.”

“The UO can better support student needs by encouraging students to seek out organizations that can support them. I think community is very important in college, and sometimes it seems like students do not realize how valuable it can be to join a club or group of people who are similar to them or can help support them.”

“Create a higher sense of community. I feel all the communities are randomly advertised or just not really shared with students.”

“Support queer/trans students and students of color better as an administration, rather than just relying on students to do it. Stop over advertising students of color to make UO seem more diverse than it actually is.”

“More social clubs or more free events for fulfillment/enrichment. More posters around campus of said free events and social clubs. The calander, as well as Handshake, feel so far removed from basic campus utilities.”

“I think there are lots of measures that can be taken to emphasize the existing support and make people more aware of their resources.”

“Make the resources that are available more known and accessible.

“Make first-year students very well known on all the resources you offer.”

“Make sure non-trad and transfer students know about the available resources. I find the people that need it are the ones that don’t hear about it.”

“More resources for undocumented students (not DACA, like undocumented). There are absolutely little to none.”

“When I went to the AEC, the people there were very kind and sweetm but the services that the AEC would accommodate for me were not very helpful or applicable to my difficulties.”

“Make it easier to work with the AEC.”

“Make the AEC more accessible.”

“More leniency. Policies can be extremely limiting and unapologetic towards personal situations.”

“Get rid of the attendance policy. I should not be penalized for missing a few days for my mental or physical health. If i am paying for my education, I should be in control of how that looks for me.”

“Stop pushing attendance policies that force people to come into lectures sick and spread it to everyone else.”

Note: Quotes were silently edited for clarity.

2023

Hierarchy of Needs

In spring 2022, we introduced a general open-ended question: How can the UO better support your basic needs (food, housing, safety, belonging, self-esteem, fulfillment)? In response, tudents had the most to say about physiological needs. A total of 99 students (24.5%) explicitly mentioned the word “food” 125 times. A total of 51 students (12.6%) explicitly mentioned the word “housing” 64 times. These frequencies underestimate the extent to which students are writing about housing or food issues, given that many other words can be and are used to describe them. Still, “food” is the second most frequently used word and “housing” is the fourth most frequently used word, overall, which puts both domains of issues top-of-mind.

The problem is that many students struggle to afford food or housing. A large portion of students (47.3%) mentioned financially related terms (394 instances). About one out of six of them (14.7) mentioned the terms “pay,” “paid,” or “wage.” About one in five of those (25.0%) specifically mentioned unionization efforts.

Students mentioned all other domains of Maslow’s hierarchy but much less frequently. A total of 21 students (5.2%) mentioned some version of the word “safe” 22 times. A total of 16 students (4.0%) mentioned some version of the word “belong” 17 times. A total of 8 students (2.0%) mentioned the word “esteem” 11 times. A total of 7 students (1.7%) mentioned some version of the word “fulfill” 10 times.

Note: The “Physiological” category combines mentions of the terms “food” and “housing.” Percentages were computed dividing the number of students who mentioned a given term (or terms in the case of “Physiological”) by the number of responses used in analysis and multiplying by 100.

Content Analysis

Themes

Overview

A group of students (n = 46) felt that their basic needs were being supported well by UO, acknowledged their privilege that their needs were otherwise satisfied, or felt that their own need satisfaction was their personal responsibility and not the university’s. However, a staggering number of students were struggling with meeting their needs or saw room for improvement in the university’s support systems (n = 404, 50.8% of the sample). Students were candid and shared how they thought the UO could improve, which we summarized in six categories of themes, described below. Themes can be thought of as both conclusions we drew from what students told us and recommendations to staff and faculty for action.

Access to Inclusive and Nutritious Food

This category focuses on ensuring all students have access to inclusive and nutritious food, which is crucial for their overall wellbeing and academic success. This means providing affordable and healthier food options that are culturally inclusive and account for dietary needs (e.g., dairy-free, gluten-free, vegan, vegetarian) in campus dining halls and venues, establishing an on-campus food pantry for students, increasing produce drops, and implementing discounted dining plans for student employees and those with dietary restrictions. Additionally, accepting SNAP benefits (i.e., food stamps) on campus and offering more meal points can significantly alleviate food insecurity among students. Educational initiatives like cooking classes and nutrition workshops can empower students to make healthier choices and manage their food budgets effectively.

Housing Affordability

This category addresses the financial burden of housing for students in Eugene, where rising rents make finding affordable accommodations difficult. Lowering on-campus housing costs and offering more scholarships can help. UO can work with local authorities and landlords to develop affordable options and secure reduced rent agreements. A dedicated housing office can assist with leases, tenant rights, and housing searches would be ideal. UO can implement stricter security measures, such as increased lighting, regular patrols, and secure access systems by campus housing. Addressing hygiene issues in on-campus housing should be a top priority, with students experiencing illnesses due to food poisoning. Guaranteeing that housing is safe, accessible, and inclusive for all students, including those with disabilities, will create a supportive and secure living environment conducive to academic and personal success.

College Affordability

This category encompasses all the issues related to the financial burdens faced by students, including high tuition and inadequate student employment opportunities. It includes scholarships, grants, financial aid programs, basic needs support, affordable tuition initiatives, and support services that help students manage and reduce the cost of college. The focus is on making higher education accessible to all students, regardless of their economic background, by providing financial resources, counseling, and tools to help them navigate college expenses. Efforts within this category aim to reduce student debt, provide transparent information about costs, and promote financial literacy to ensure students can afford and complete their education without undue financial burden. Addressing the high cost of tuition and other related expenses is essential to make college more affordable for all students to include transportation costs. Lowering tuition rates for both in-state and out-of-state students can significantly reduce financial stress. Increasing student wages and providing more financial aid and scholarships will help cover the various costs associated with college. Offering grants and financial support for expenses beyond tuition, such as books and supplies, can further alleviate the financial burden on students. Utilizing donations from alumni and boosters to support non-athlete students can provide additional financial resources and opportunities for those in need.

Safety

This category focuses on ensuring the physical safety and security of students on and off campus. It includes concerns and suggestions related to campus lighting, security measures, transportation safety, and overall campus environment. The goal is to create a safe and secure atmosphere where students feel protected and supported, both during the day and at night. Students indicate the importance of adequate security personnel, effective safety protocols, and accessible transportation options to enhance students’ sense of safety and wellbeing. Expanding Duck Rides to operate during all hours of darkness can enhance safety. Protecting the safety of students on and around campus can foster a secure and supportive educational environment. Additionally, enhancing emergency response protocols and providing clear communication about safety resources can address concerns about violence, theft, and other dangers. Creating safe housing options near campus for those who cannot afford to live on campus, particularly for marginalized students, and promoting a culture of respect and inclusion assure that all students feel safe and supported in their daily lives.

Mental Health and Wellbeing

This category addresses the mental health challenges students face, including the availability and accessibility of counseling services, managing academic stress, and ensuring physical health. It emphasizes the need for comprehensive mental health support and effective safety measures. Normalizing mental health issues and reducing their stigma are essential for fostering a supportive environment. Flexible attendance policies can accommodate students dealing with mental health struggles or illness. Expanding mental health services and support systems is a component of improving student wellbeing. Removing the cap on therapy sessions and providing information about local therapists and group therapy options will ensure continuous access to mental health care. Increasing the number of counselors and improving the accessibility of mental health resources can meet the growing demand for these services. Establishing regular mental health check-ins, offering emergency mental health support, and providing mental health days can help students manage stress and maintain their mental health. Educating professors on how to support students with mental health issues will create a more understanding and compassionate academic environment.

Community and Inclusion

This category addresses the social and inclusion-related challenges students face, such as difficulties in social integration, lack of support for marginalized and neurodiverse students, and the need for a more inclusive campus environment. It emphasizes the importance of creating a welcoming and supportive community for all students, that everyone, regardless of their background, has access to the resources and support needed to succeed. An approach to inclusion encompasses services, programs, and initiatives aimed at academic support, mental health services, cultural and identity-based resources, and accommodations for students with disabilities. Supporting marginalized students and promoting equity can help create a sense of belonging. Providing tailored resources and programs for students with children, nontraditional students, commuter students and those with disabilities helps these students feel supported, included, and can help them connect with their peers.

Academic Support and Resources

This category focuses on the academic challenges students encounter, emphasizing the need for improved tutoring, advising, and course flexibility. A peer-to-peer model can enhance learning experiences, while offering flexible makeup work and assignment deadlines can accommodate diverse student needs. Enhancing the awareness and accessibility of academic support services is vital for creating a supportive learning environment. Fostering a culture of understanding and compassion among faculty, along with revising attendance policies to accommodate student needs, can create a more inclusive academic environment. Potential changes to the term system can provide students with the adaptability they need to thrive. Establishing a one-stop hub, whether in-person or online, for all student support services can streamline the process of seeking assistance. Providing additional tutoring and study resources, especially for challenging subjects, will help students overcome academic obstacles and achieve their educational goals.

Frequencies

Theme	Description	n	%
Access to Inclusive and Nutritious Food	This category focuses on ensuring all students have access to inclusive and nutritious food, which is crucial for their overall wellbeing and academic success. This means providing affordable and healthier food options that are culturally inclusive and account for dietary needs (e.g., dairy-free, gluten-free, vegan, vegetarian) in campus dining halls and venues, establishing an on-campus food pantry for students, increasing produce drops, and implementing discounted dining plans for student employees and those with dietary restrictions. Additionally, accepting SNAP benefits (i.e., food stamps) on campus and offering more meal points can significantly alleviate food insecurity among students. Educational initiatives like cooking classes and nutrition workshops can empower students to make healthier choices and manage their food budgets effectively.	146	36.1

Housing Affordability	This category addresses the financial burden of housing for students in Eugene, where rising rents make finding affordable accommodations difficult. Lowering on-campus housing costs and offering more scholarships can help, while partnering with local authorities and landlords can secure reduced rent agreements. Implementing stricter security measures, addressing hygiene issues, and ensuring housing is safe, accessible, and inclusive will create a supportive environment conducive to student success.	62	15.3

College Affordability	This category addresses the financial burdens faced by students, including high tuition and inadequate employment opportunities. It covers scholarships, grants, financial aid programs, and support services to help manage and reduce college costs. The goal is to make higher education accessible to all by providing financial resources, counseling, and tools to navigate expenses, ultimately reducing student debt. Efforts include lowering tuition rates, increasing student wages, and offering additional financial support for expenses like books and supplies.	143	35.4

Safety	This category focuses on ensuring the physical safety and security of students on and off campus, addressing concerns related to campus lighting, security measures, transportation safety, and overall campus environment. The goal is to create a secure atmosphere where students feel protected and supported at all times. Enhancing emergency response protocols, expanding Duck Rides to operate during all hours of darkness, and promoting a culture of respect and inclusion are essential for fostering a safe and supportive educational environment.	35	8.7

Mental Health and Wellbeing	This category addresses the mental health challenges students face, focusing on the availability and accessibility of counseling services, managing academic stress, and ensuring physical health. It emphasizes comprehensive mental health support, normalizing mental health issues, and reducing stigma. Flexible attendance policies and expanding mental health services are essential for student wellbeing. Increasing the number of counselors, removing caps on therapy sessions, and providing regular mental health check-ins and emergency support can help students manage stress and maintain their mental health.	62	15.3

Community and Inclusion	This category addresses social and inclusion-related challenges, such as difficulties in social integration and lack of support for marginalized and neurodiverse students. It emphasizes creating a welcoming and supportive community where all students have access to necessary resources. This includes academic support, mental health services, cultural and identity-based resources, and accommodations for students with disabilities. Tailored programs for students with children, nontraditional students, commuters, and those with disabilities help foster a sense of belonging and connection among peers.	92	22.8

Academic Support and Resources	This category addresses academic challenges, emphasizing the need for improved tutoring, advising, and course flexibility. A peer-to-peer model, flexible makeup work, and assignment deadlines can accommodate diverse student needs. Enhancing the accessibility and awareness of academic support services and fostering a compassionate faculty culture can create a more inclusive environment. Establishing a one-stop hub for all student support services can streamline assistance and help students overcome academic obstacles.	80	19.8

Note: Themes can be thought of as both conclusions we drew from what students told us and recommendations to staff and faculty for action. Responses were open-coded using Grounded Theory (Strauss & Corbin, 2006; Tracy, 2019), which is a framework used to understand people’s experiences by building on themes from qualitative data they provide. Quotes were silently edited for clarity.

Wordnet

There are some interesting threads in the 2-to-3-term clusters of the network depiction of adjacent adjectives and nouns below, but the main stories are in the big cluster on the left side and in the 5-term string below it. Students are concerned about food security and want more options and resources, including free or cheaper food and a better pantry situation. Students are also concerned about rising housing costs and want more affordable, accessible housing options. They also want lower tuition and more support through and better access to financial and mental health resources and opportunities for social interaction.

Note: Wordnet = adjacent ($\le$ 1 word apart) adjective-noun bigrams visualized as a network. Line thickness represents frequency of cooccurrence. The criterion for inclusion of bigram in the wordnet was cooccurrence $\gt$ 3 (min = 4, max = 21). If a word seems positively valenced, it may have been negated (e.g., preceded by the word “not”). Text data were annotated using the udpipe package.

Wordcloud

Keywords extracted and displayed in the wordcloud below largely confirm the descriptions in the Hierarchy of Needs, Themes, and Wordnet sections. Students are struggling to make ends meet and want more tangible support of their physiological needs, particularly food and housing.

Note: Word size represents frequency of keyword occurrence. Hover over words to see frequencies. The criterion for inclusion of a keyword in the wordcloud was occurrence $\gt$ 3. If a word seems positively valenced, it may have been negated (e.g., preceded by the word “not”). Text data were annotated using the udpipe package. Keywords were extracted using the Rapid Keyword Extraction (RAKE) algorithm (Rose et al., 2010).

Quotes

“I think allowing more food options for people with UO meal plans and lowering the cost of simple food items would help extremely.”

“Have more fruit and vegetable options in the dining halls or lower the number of points they cost so you can get more of them for your room.”

“There needs to be more options for those with food allergies or sensitivities that are not super expensive or cost a lot of points.”

“It would be extremely helpful if all dining halls were open until 9pm. Just that one extra hour would benefit many students because people have work, classes, and other commitments that stop them from having a well-rounded dinner.”

“I found myself running out of meal points during every single finals week. I had to use my limited real money to pay for food.”

“Food resources for my peers. Accept SNAP in the EMU, construction of a new food pantry that is not in a church, food recovery program for dining hall food waste.”

“Bring back Ducks Feeding Ducks being 3 times per term.”

“The price of a food plan is outrageous and the stores on campus are so overpriced. So, students without means of transportation have to pay more. Given how much we pay for tuition and for everything else at the UO, it should be cheaper for students, not more expensive.”

“It is crazy how expensive everything is. As a full-time student, it is hard to afford college itself, let alone everything else that has to be bought along with it.”

“I know this will never be listened to, but there is no moral argument for tuition being as high as it is, especially considering the pay that is now promised to college grads is basically still poverty wages. Please stop pretending to care. Lower tuition.”

“I think housing and food should cost less on the UO campus. A lot of students are struggling to survive on their own without the support of their parents. So, making housing and dining halls more financially accessible would benefit everyone.”

“Housing is extremely expensive in Eugene, and tuition continues to get more expensive. I highly doubt the university will do anything about this, but paying my tuition and housing is a big part of why I could not afford enough food some months.”

“Cut the price. If I can get a fully furnished apartment for $900 a month in Eugene then there is no reason for dorms to be this expensive. Dorms are supposed to be the cheap option. If three people are paying $1,500 a month on a room, then that room better be worth $4,500 a month.”

“Raise the estimated cost of attendance to accommodate realistic living expenses for nontraditional students. This is typically about $1,500 to $2,000 per month when on a budget. Gas, insurance, and car cost $500 per month. Owning a car is not a lifestyle choice; it a necessity for anyone who has to work or has family members to support. Rent for a single bedroom is usually more than $1,000 per month here. Utilities cost $200 per month in the cold months. Food is at the very minimum $300 per month. So, a total of $24,000 per year is a very modest estimate and assumes zero extra expenses: technology, clothes, other essentials. So, why is the estimated cost just under $19,000? Living on that amount of money for the year is only possible for mostly white students who are supported by family in some way. Please just raise this estimate so we can feed ourselves while here.”

“I think that safety at this school is non-existent. I tend to feel endangered when I have classes at night and have to leave campus around 8:00pm.”

“Improve safety when walking from campus to Riley and vice versa. Maybe add more street lights, emergency buttons, and more signs.”

“I think there needs to be more blue light systems throughout campus, specifically near the Knight Library. Often there are parties on the other side of campus, and late at night you have to walk back through that quad area of the Knight Library. Although it is lit, the closest blue light emergency button is about 100 yards away, and I am not going to be able to sprint to it if I feel unsafe. I think this needs to be addressed.”

“Increase the number of counseling sessions at the Counseling Center or allow counselors to accept outside insurance.”

“Having more than just six free counseling sessions per term would be super helpful to students who need weekly services and can not afford or access off campus services.”

“A long term option for counseling would be great. I have utilized the counseling services but all they could recommend was other offices for long term aid. I have since been on a waiting list for a few months in an attempt to get help but have had no luck with outside providers.”

“Some professors have had a hard time understanding that I need to take days to look after my mental health and are pretty unforgiving with deadlines and attendance policies.”

“It’s significant when professors address and act towards the mental and physical wellbeing of students. So, continued support to give them awareness and tools to help makes a positive difference.”

“The UO needs to ensure that professors and GEs have the capacity to give their students the attention they deserve. I have had numerous professors that were incredibly helpful but others have struggled to understand situations of duress I was under, and as a result my academic performance suffered greatly.”

“Better educate people on where to find the resources.”

“Tell students that you care and actually mean it.”

“The reason the absence policy is disastrous for nontraditional students is that it makes it impossible for teachers to work with students who have extenuating circumstances. The policy makes it feel like the UO is geared toward kids straight out of high school who have their parents paying for everything.”

“There should be more services for students as parents. All organizations should have events at more varied or for longer lasting times. I often cannot attend things I want to because of children, work, or school duties.”

“I think there should be more inclusive spaces and programming for nontraditional students. I did not get the first-year experience because it was tailored toward traditional students.”

“It would be nice if there was some way for socially anxious, awkward people like myself to meet others like them. I have found that socially, anxious, awkward people tend to get along better with people like themselves rather than more outgoing people.”

Note: Quotes were silently edited for clarity.

2022

Hierarchy of Needs

In spring 2022, we introduced a general open-ended question: How can the UO better support your basic needs (food, housing, safety, belonging, self-esteem, fulfillment)? In response, students had the most to say about physiological needs. A total of 151 students (23.4%) explicitly mentioned the word “food” 193 times. A total of 56 students (8.7%) explicitly mentioned the word “housing” 66 times. These frequencies underestimate the extent to which students are writing about housing or food issues, given that many other words can be and are used to describe housing of food issues. Still, “food” is the second most frequently used word and “housing” is the fourth most frequently used word, overall, which puts both domains of issues top-of-mind.

The problem is that many students struggle to afford food or housing. A large portion (45.3%) of students mentioned financially related terms (541 instances), with 15.8% of them mentioning the more specific terms “pay,” “paid,” or “wage.”

Students mentioned all other domains of Maslow’s hierarchy but much less frequently. A total of 40 students (6.2%) mentioned some version of the word “safe” 57 times. A total of 16 students (2.5%) mentioned some version of the word “belong” 17 times. A total of 9 students (1.4%) mentioned the word “esteem” 9 times. A total of 7 students (1.1%) mentioned some version of the word “fulfill” 7 times.

Note: The “Physiological” category combines mentions of the terms “food” and “housing.” Percentages were computed dividing the number of students who mentioned a given term (or terms in the case of “Physiological”) by the number of responses used in analysis and multiplying by 100.

Content Analysis

Themes

Overview

A group of students (n = 63) felt that their basic needs were being supported well by UO, acknowledged their privilege that their needs were otherwise satisfied, or felt that their own need satisfaction was their personal responsibility and not the university’s. However, a staggering number of students were struggling with meeting their needs or saw room for improvement in the university’s support systems (n = 644, 54.8% of the sample). Students were candid and shared how they thought the UO could improve, which we summarized in six categories of themes, described below. Themes can be thought of as both conclusions we drew from what students told us and recommendations to staff and faculty for action.

Access to Inclusive and Nutritious Food

This category focuses on ensuring all students have access to inclusive and nutritious food, which is crucial for their overall wellbeing and academic success. This means providing affordable and healthier food options that are culturally inclusive and account for dietary needs (e.g., dairy-free, gluten-free, vegan, vegetarian) in campus dining halls and venues, establishing an on-campus food pantry for students, increasing produce drops, and implementing discounted dining plans for student employees and those with dietary restrictions. Additionally, accepting SNAP benefits (i.e., food stamps) on campus and offering more meal points can significantly alleviate food insecurity among students. Educational initiatives like cooking classes and nutrition workshops can empower students to make healthier choices and manage their food budgets effectively.

Housing Affordability

This category addresses the financial burden of housing for students in Eugene, where rising rents make finding affordable accommodations difficult. Lowering on-campus housing costs and offering more scholarships can help. UO can work with local authorities and landlords to develop affordable options and secure reduced rent agreements. A dedicated housing office can assist with leases, tenant rights, and housing searches would be ideal. UO can implement stricter security measures, such as increased lighting, regular patrols, and secure access systems by campus housing. Addressing hygiene issues in on-campus housing should be a top priority, with students experiencing illnesses due to food poisoning. Guaranteeing that housing is safe, accessible, and inclusive for all students, including those with disabilities, will create a supportive and secure living environment conducive to academic and personal success.

College Affordability

This category encompasses all the issues related to the financial burdens faced by students, including high tuition and inadequate student employment opportunities. It includes scholarships, grants, financial aid programs, basic needs support, affordable tuition initiatives, and support services that help students manage and reduce the cost of college. The focus is on making higher education accessible to all students, regardless of their economic background, by providing financial resources, counseling, and tools to help them navigate college expenses. Efforts within this category aim to reduce student debt, provide transparent information about costs, and promote financial literacy to ensure students can afford and complete their education without undue financial burden. Addressing the high cost of tuition and other related expenses is essential to make college more affordable for all students to include transportation costs. Lowering tuition rates for both in-state and out-of-state students can significantly reduce financial stress. Increasing student wages and providing more financial aid and scholarships will help cover the various costs associated with college. Offering grants and financial support for expenses beyond tuition, such as books and supplies, can further alleviate the financial burden on students. Utilizing donations from alumni and boosters to support non-athlete students can provide additional financial resources and opportunities for those in need.

Safety

This category focuses on ensuring the physical safety and security of students on and off campus. It includes concerns and suggestions related to campus lighting, security measures, transportation safety, and overall campus environment. The goal is to create a safe and secure atmosphere where students feel protected and supported, both during the day and at night. Students indicate the importance of adequate security personnel, effective safety protocols, and accessible transportation options to enhance students’ sense of safety and wellbeing. Expanding Duck Rides to operate during all hours of darkness can enhance safety. Protecting the safety of students on and around campus can foster a secure and supportive educational environment. Additionally, enhancing emergency response protocols and providing clear communication about safety resources can address concerns about violence, theft, and other dangers. Creating safe housing options near campus for those who cannot afford to live on campus, particularly for marginalized students, and promoting a culture of respect and inclusion assure that all students feel safe and supported in their daily lives.

Mental Health and Wellbeing

This category addresses the mental health challenges students face, including the availability and accessibility of counseling services, managing academic stress, and ensuring physical health. It emphasizes the need for comprehensive mental health support and effective safety measures. Normalizing mental health issues and reducing their stigma are essential for fostering a supportive environment. Flexible attendance policies can accommodate students dealing with mental health struggles or illness. Expanding mental health services and support systems is a component of improving student wellbeing. Removing the cap on therapy sessions and providing information about local therapists and group therapy options will ensure continuous access to mental health care. Increasing the number of counselors and improving the accessibility of mental health resources can meet the growing demand for these services. Establishing regular mental health check-ins, offering emergency mental health support, and providing mental health days can help students manage stress and maintain their mental health. Educating professors on how to support students with mental health issues will create a more understanding and compassionate academic environment.

Community and Inclusion

This category addresses the social and inclusion-related challenges students face, such as difficulties in social integration, lack of support for marginalized and neurodiverse students, and the need for a more inclusive campus environment. It emphasizes the importance of creating a welcoming and supportive community for all students, that everyone, regardless of their background, has access to the resources and support needed to succeed. An approach to inclusion encompasses services, programs, and initiatives aimed at academic support, mental health services, cultural and identity-based resources, and accommodations for students with disabilities. Supporting marginalized students and promoting equity can help create a sense of belonging. Providing tailored resources and programs for students with children, nontraditional students, commuter students and those with disabilities helps these students feel supported, included, and can help them connect with their peers.

Academic Support and Resources

This category focuses on the academic challenges students encounter, emphasizing the need for improved tutoring, advising, and course flexibility. A peer-to-peer model can enhance learning experiences, while offering flexible makeup work and assignment deadlines can accommodate diverse student needs. Enhancing the awareness and accessibility of academic support services is vital for creating a supportive learning environment. Fostering a culture of understanding and compassion among faculty, along with revising attendance policies to accommodate student needs, can create a more inclusive academic environment. Potential changes to the term system can provide students with the adaptability they need to thrive. Establishing a one-stop hub, whether in-person or online, for all student support services can streamline the process of seeking assistance. Providing additional tutoring and study resources, especially for challenging subjects, will help students overcome academic obstacles and achieve their educational goals.

Frequencies

Theme	Description	n	%
Access to Inclusive and Nutritious Food	This category focuses on ensuring all students have access to inclusive and nutritious food, which is crucial for their overall wellbeing and academic success. This means providing affordable and healthier food options that are culturally inclusive and account for dietary needs (e.g., dairy-free, gluten-free, vegan, vegetarian) in campus dining halls and venues, establishing an on-campus food pantry for students, increasing produce drops, and implementing discounted dining plans for student employees and those with dietary restrictions. Additionally, accepting SNAP benefits (i.e., food stamps) on campus and offering more meal points can significantly alleviate food insecurity among students. Educational initiatives like cooking classes and nutrition workshops can empower students to make healthier choices and manage their food budgets effectively.	210	32.6

Housing Affordability	This category addresses the financial burden of housing for students in Eugene, where rising rents make finding affordable accommodations difficult. Lowering on-campus housing costs and offering more scholarships can help, while partnering with local authorities and landlords can secure reduced rent agreements. Implementing stricter security measures, addressing hygiene issues, and ensuring housing is safe, accessible, and inclusive will create a supportive environment conducive to student success.	66	10.2

College Affordability	This category addresses the financial burdens faced by students, including high tuition and inadequate employment opportunities. It covers scholarships, grants, financial aid programs, and support services to help manage and reduce college costs. The goal is to make higher education accessible to all by providing financial resources, counseling, and tools to navigate expenses, ultimately reducing student debt. Efforts include lowering tuition rates, increasing student wages, and offering additional financial support for expenses like books and supplies.	173	26.9

Safety	This category focuses on ensuring the physical safety and security of students on and off campus, addressing concerns related to campus lighting, security measures, transportation safety, and overall campus environment. The goal is to create a secure atmosphere where students feel protected and supported at all times. Enhancing emergency response protocols, expanding Duck Rides to operate during all hours of darkness, and promoting a culture of respect and inclusion are essential for fostering a safe and supportive educational environment.	64	9.9

Mental Health and Wellbeing	This category addresses the mental health challenges students face, focusing on the availability and accessibility of counseling services, managing academic stress, and ensuring physical health. It emphasizes comprehensive mental health support, normalizing mental health issues, and reducing stigma. Flexible attendance policies and expanding mental health services are essential for student wellbeing. Increasing the number of counselors, removing caps on therapy sessions, and providing regular mental health check-ins and emergency support can help students manage stress and maintain their mental health.	125	19.4

Community and Inclusion	This category addresses social and inclusion-related challenges, such as difficulties in social integration and lack of support for marginalized and neurodiverse students. It emphasizes creating a welcoming and supportive community where all students have access to necessary resources. This includes academic support, mental health services, cultural and identity-based resources, and accommodations for students with disabilities. Tailored programs for students with children, nontraditional students, commuters, and those with disabilities help foster a sense of belonging and connection among peers.	125	19.4

Academic Support and Resources	This category addresses academic challenges, emphasizing the need for improved tutoring, advising, and course flexibility. A peer-to-peer model, flexible makeup work, and assignment deadlines can accommodate diverse student needs. Enhancing the accessibility and awareness of academic support services and fostering a compassionate faculty culture can create a more inclusive environment. Establishing a one-stop hub for all student support services can streamline assistance and help students overcome academic obstacles.	116	18.0

Note: Themes can be thought of as both conclusions we drew from what students told us and recommendations to staff and faculty for action. Responses were open-coded using Grounded Theory (Strauss & Corbin, 2006; Tracy, 2019), which is a framework used to understand people’s experiences by building on themes from qualitative data they provide. Quotes were silently edited for clarity.

Wordnet

There are some interesting threads in the 2-to-3-term clusters of the network depiction below of adjacent adjectives and nouns, but the main story is in the big cluster with “food,” “more,” and “health” centroids. Students are concerned about food security and want more options and resources, including free or cheaper food, more nutritious food, and a better pantry situation. Students also want cheaper housing, lower tuition, and more support through and better access to financial and mental health resources and opportunities for social interaction.

Note: Wordnet = adjacent ($\le$ 1 word apart) adjective-noun bigrams visualized as a network. Line thickness represents frequency of cooccurrence. The criterion for inclusion of bigram in the wordnet was cooccurrence $\gt$ 3 (min = 4, max = 61). If a word seems positively valenced, it may have been negated (e.g., preceded by the word “not”). Text data were annotated using the udpipe package.

Wordcloud

Keywords extracted and displayed in the wordcloud below largely confirm the descriptions in the Hierarchy of Needs, Themes, and Wordnet sections. Students are struggling to make ends meet and want more tangible support of their physiological needs (food and housing).

Note: Word size represents frequency of keyword occurrence. Hover over words to see frequencies. The criterion for inclusion of a keyword in the wordcloud was occurrence $\gt$ 3. If a word seems positively valenced, it may have been negated (e.g., preceded by the word “not”). Text data were annotated using the udpipe package. Keywords were extracted using the Rapid Keyword Extraction (RAKE) algorithm (Rose et al., 2010).

Quotes

“I definitely feel like food is a good way to help students out. When you are so consumed with where your next meal is going to come from, it can be hard to focus on school, social engagement, or self-fulfillment.”

“It would be nice to have more food variety and more kitchens so I can cook for myself.”

“A larger variety of food. I have a food allergy and have very few options. Most days, I end up eating the same few meals.”

“Having more services for those with food insecurities such as food banks, monthly vouchers, etc. would help.”

“It would be beneficial if there was a program where food pantries would exist on campus because I do not have transportation to travel elsewhere.”

“Take food stamps (SNAP EBT) on campus. It’s very discriminatory that no places on campus take them.”

“I think having more food and housing security resources would be very helpful. My friends and I struggled at times with having enough food and we solely relied on food stamps. If it was not for this resource, I would not be eating enough.”

“The meal plans suck. We are literally being robbed every week. If we do not spend all of the points on the meal plan, it would make more sense to have anything remaining at the end of the term credited back to us.”

“Discounted dining hall food for student employees.”

“I live with family but if the housing cost was low, I would love to live on campus to be more involved.”

“Providing cheaper housing and more scholarships or stipends can alleviate the financial burden and enable more students to participate fully in campus life.”

“The University of Oregon needs to provide more financial resources to offset the cost of attendance and reduce the cost of student housing.”

“Stop charging excessive costs to even go to school here… I pay almost 60k to go to school here and last year over 50% of students asked for rental or food assistance.”

“UO security absolutely needs to step up and make sure that all campus buildings and spaces are safe from a physical standpoint.”

“It is a bit concerning that in the past three terms my dorm has had a man break in and hold students hostage, an arsonist and various fires…the lack of an elevator in my building is incredibly questionable from a legal standpoint.”

“I would appreciate extended Duck Ride hours during the winter time and for it to begin as soon as it gets dark. I had to walk home in the dark at 5pm because Duck Rides did not start until 6pm.”

“Safety-wise, campus needs to be more secure. There are so many people who live off campus who have experienced very dangerous situations in walking to and from school.”

“We need better access to counseling. Right now, it’s very restricted.”

“The counseling services need to be expanded upon. I cannot get into the university health center counseling unless I am in a crisis, and I cannot afford preventative care.”

“Making counseling services longer than the 6-session limit and providing more mental health support would be super helpful to students who need weekly services and cannot afford off-campus services.”

Better advertising and dissemination of the available academic support resources would ensure that students are well-informed and can easily access the help they need.”

“Create a main hub, like a campus-wide app, to find all available UO resources. Continue to create community and spaces that bring common people together.”

“Get rid of the stupid quarter system so I don’t feel stressed all the time and have more time to get to know my peers and to participate in social events and clubs.”

“Creating a community where others are extremely kind would spread feelings of belonging, non-judgment, acceptance, and compassion, starting with staff and student representatives”

“Have more neurodiverse clubs and improve the AEC. I have found very few social things related to neurodiverse issues, and the AEC needs improvement. A few of my friends have had bad experiences there, and they do not help, making situations worse.”

Note: Quotes were silently edited for clarity.

Recommendations

Based on the multimethod content analysis of students’ responses to the open-ended item, we have developed the following recommendations:

Access to Inclusive and Nutritious Food

Expand Nutritious Dining Options: Increase the availability of affordable and healthy food choices in campus dining halls, including more options for students with dietary restrictions such as gluten-free, dairy-free, vegan, vegetarian, and ketogenic diets. Extend dining hall hours to accommodate students’ varied schedules.
On-Campus Food Pantry: Prioritize the establishment and funding of an on-campus food pantry to support students facing food insecurity. Ensure it is accessible and well publicized.
On-Campus Produce Drops: Increase the frequency of produce drops to more than once per week at a different time of day.
Discounted Dining Plans: Implement discounted dining plans for student employees and those with dietary restrictions to make meal plans more affordable.
Ducks Feeding Ducks: Increase the frequency of Ducks Feeding Ducks from once per term to three times or more per term. During annual fundraising efforts, advertise the Ducks Feeding Ducks for alumni, faculty, and staff to contribute to.
Accept SNAP EBT on Campus: Allow the use of SNAP EBT (i.e., food stamps) at campus cafes, markets, and dining facilities to support low-income students.
Meal Points Rollover: Allow students to rollover unused meal points to subsequent terms or convert them into Duck Bucks to avoid wastage.
Cooking and Nutrition Workshops: Offer cooking classes and nutrition workshops to help students make healthier food choices and manage their food budgets effectively.

Housing Affordability

Lower On-Campus Housing Costs: Reduce the cost of on-campus housing through subsidies or by reallocating funds to make living on campus more affordable.
Realistic Cost-of-Living Estimates: Work with state and federal partners to advocate for updated cost-of-living estimates for the Eugene area. Current estimates are too low, which results in insufficient aid packages for students.
Affordable Housing Partnerships: Partner with local landlords to secure reduced rent agreements for students and provide resources to help them navigate the off-campus housing market.
Dedicated Housing Office: Create a housing office that assists students with lease agreements, tenant rights, and housing searches.
Safety in Housing: Implement stricter security measures in dorms, such as increased lighting, regular security patrols, and secure access systems. Address hygiene issues promptly to ensure safe living conditions.

College Affordability

Lower Tuition Rates: Advocate for lower tuition rates for both in-state and out-of-state students to reduce financial stress.
Increase Financial Aid and Scholarships: Provide more financial aid, scholarships, and grants to help cover the various costs associated with college, including books and supplies.
Student Employment Opportunities: Increase on-campus job opportunities and raise student wages to help cover living expenses.
Utilize Alumni Donations: Use donations from alumni and boosters to support non-athlete students and provide additional financial resources for those in need.

Safety

Improve Campus Lighting: Increase lighting in poorly lit areas around campus to enhance visibility and safety at night.
Enhance Security Measures: Ensure adequate security personnel and effective safety protocols, including regular safety patrols and secure access to buildings.
Expand Transportation Services: Extend the operating hours of services like Duck Rides to all hours of darkness to ensure safe transportation options for students.
Emergency Response Protocols: Enhance emergency response protocols and provide clear communication about safety resources and procedures.

Mental Health and Wellbeing

Expand Counseling Services: Increase the number of counselors and extend the session limits to ensure continuous mental health support for students.
Flexible Attendance Policies: Implement flexible attendance policies to accommodate students dealing with mental health struggles or illness.
Regular Mental Health Check-ins: Establish regular mental health check-ins and offer emergency mental health support.
Educate Faculty: Train professors on how to support students with mental health issues, including ways to speak about stress and mental health inclusively.
Increase Accessibility: Improve the accessibility of mental health resources and provide clear information on local therapists and group therapy options.

Community and Inclusion

Support for Marginalized and Neurodiverse Students: Provide tailored resources and programs for marginalized and neurodiverse students, including more neurodiverse clubs and improved services at the Accessible Education Center (AEC).
Promote Diversity and Inclusion Initiatives: Increase diversity and inclusion initiatives with balanced messaging to create a welcoming campus environment.
Social Engagement Opportunities: Increase opportunities for social engagement and community building, particularly for commuter and nontraditional students, to help them connect with their peers.
Inclusive Campus Environment: Foster a culture of understanding and compassion among students and staff, promoting a sense of belonging and support for all students.

Academic Support and Resources

Improve Tutoring Services: Expand tutoring services, especially for challenging subjects like computer science and math, to ensure adequate academic support.
Flexible Course Policies: Offer flexible makeup work and assignment deadlines to accommodate diverse student needs.
Advertise Support Services: Increase the advertisement of academic support services to ensure students are well informed about the help available to them.
One-Stop Support Hub: Establish a centralized hub, either in-person or online, for all student support services to include basic needs to streamline the process of seeking assistance.
Faculty Accountability: Hold faculty accountable for their engagement with students and ensure they are accessible for academic support.

Supplemental Method

Sources, Sizes, and Characteristics

SWaSI end-of-year check-in data collection occurs in the latter half (weeks 6 through 10) of every spring term. Food security and wellbeing data came from the spring 2024, 2023, 2022, and 2021 waves of that. Measurements of core constructs (ability uncertainty, belonging uncertainty, loneliness, social support, stereotype threat, general health, life satisfaction, stress, self-assurance, sadness) and other constructs (food security, place-based belonging, etc.) were block randomized in presentation. Demographics data (gender identity, sexuality, etc.) were collected at the end.

Recruitment procedures shifted in spring 2022. Instead of following up with participating students longitudinally by emailing them across the second half of each spring term – this is why 2021 analyses excluded fourth-year students – the end-of-year check-in effectively became cross-sectional and various campus units and individual colleagues promoted the survey through various communications media. By collecting identifiers, we are still able to cobble together multiple data sources and time points.

SWaSI baseline data collection happens every summer prior to matriculation. Some demographics data from baseline were used here. Global socioeconomic status, family social class, and high school quality were exclusively from baseline SWaSI. Gender identity/sexuality from baseline SWaSI were used to fill missing data from end-of-year check-in. And parent education level from baseline SWaSI were used to fill missing data from admissions records.

Demographics data came from:

admissions data only (race/ethnicity, transfer status)
baseline SWaSI data only (global socioeconomic status, family social class, high school quality)
end-of-year + baseline SWaSI data (gender identity/sexuality) – missingness in given end-of-year wave is filled with baseline data
admissions data + baseline SWaSI data (parent education level) – missingness in admissions data is filled with baseline SWaSI data.

Additional food security data came from the National College Health Assessment administered in spring 2022. Additional food security data as well as housing security and homelessness data came from the Hope Center Student Basic Needs Survey administered in winter 2023.

SWaSI Data

Totals of n = 1,153 and n = 1,402 admitted, first-through-fourth-year, U.S. undergraduates who contributed at least one usable datum in spring 2024 and 2022 end-of-year check-ins, respectively, and n = 837 admitted, first-through-third-year, U.S. undergraduates who contributed at least one usable datum in spring 2021 were considered potential participants. Samples for overall estimates of food security were reduced by missing food security data. Missing data has been increasing due to increased survey noncompletion, which is likely partly due to changes in recruitment methods implemented starting in 2022 and to a syndrome known as survey fatigue.

Overall Sample Size and Percent Missing

Year	n	% Missing
2024	1,233	1.8
2023	795	31.0
2022	1,175	16.2
2021	790	5.6

Note: Table produced using the huxtable package.

After filtering out cases that had no data on food security, there were missing data in demographics and wellbeing variables that lowered sample sizes for any given subsequent analysis. By how much sample sizes were lowered depended on model and data source.

Model Sample Size

For consistency, uncontrolled and controlled models of wellbeing consequences excluded cases missing data on outcomes of interest (stress, sadness, life satisfaction), the predictor of interest (food security), and relevant control covariates.

2024

Single-predictor models using data that included demographics from only admissions data (race/ethnicity, transfer status) had an average sample size of n = 789. Single-predictor models using data that included demographics from only baseline SWaSI data (global socioeconomic status, family social class, high school quality) had an average sample size of n = NaN. Single-predictor models using data that included demographics from end-of-year + baseline SWaSI data (gender identity/sexuality primarily relied on end-of-year data and when missing was filled with baseline data when available) had an average sample size of n = 750. Single-predictor models using data that included demographics from admissions + baseline SWaSI data (parent education level primarily relied on admissions data and when missing was filled with baseline SWaSI data when available) had an average sample size of n = 586. Multiple-predictor models using data had an average sample size of n = 547. Wellbeing consequences models using data had an average sample size of n = 519.

2022

Single-predictor models using data that included demographics from only admissions data (race/ethnicity, transfer status) had an average sample size of n = 1,162. Single-predictor models using data that included demographics from only baseline SWaSI data (global socioeconomic status, family social class, high school quality) had an average sample size of n = NaN. Single-predictor models using data that included demographics from end-of-year + baseline SWaSI data (gender identity/sexuality primarily relied on end-of-year data and when missing was filled with baseline data when available) had an average sample size of n = 1,144. Single-predictor models using data that included demographics from admissions + baseline SWaSI data (parent education level primarily relied on admissions data and when missing was filled with baseline SWaSI data when available) had an average sample size of n = 812. Multiple-predictor models using data had an average sample size of n = 790. Wellbeing consequences models using data had an average sample size of n = 782.

2021

Single-predictor models using data that included demographics from only admissions data (race/ethnicity, transfer status) had an average sample size of n = 759. Single-predictor models using data that included demographics from only baseline SWaSI data (global socioeconomic status, family social class, high school quality) had an average sample size of n = NaN. Single-predictor models using data that included demographics from end-of-year + baseline SWaSI data (gender identity/sexuality primarily relied on end-of-year data and when missing was filled with baseline data when available) had an average sample size of n = 780. Single-predictor models using data that included demographics from admissions + baseline SWaSI data (parent education level primarily relied on admissions data and when missing was filled with baseline SWaSI data when available) had an average sample size of n = 665. Multiple-predictor models using data had an average sample size of n = 636. Wellbeing consequences models using data had an average sample size of n = 708.

Exact ns

Year	Model Type	Outcome	Covariate(s)	n
2024	Single-Predictor	Food Security	Gender Identity/Sexuality	1,179
			Disability/Neurodivergence	1,160
			Race/Ethnicity	1,219
			Socioeconomic Status	962
			Transfer Status	1,227
	Multiple-Predictor		GISO, DN, RE, SES, TS	918
	Wellbeing Consequences	Stress	FS, DN, GISO, TS	903
		Sadness	FS, DN, GISO, RE	878
		Life Satisfaction	FS, DN, GISO, RE, TS	890
2023	Single-Predictor	Food Security	Gender Identity/Sexuality	765
			Disability/Neurodivergence	734
			Race/Ethnicity	784
			Socioeconomic Status	586
			Transfer Status	794
	Multiple-Predictor		GISO, DN, RE, SES, TS	547
	Wellbeing Consequences	Stress	FS, DN, GISO, RE	515
		Sadness	FS, DN, GISO, RE	520
		Life Satisfaction	FS, DN, GISO, RE, TS	522
2022	Single-Predictor	Food Security	Gender Identity/Sexuality	1,144
			Race/Ethnicity	1,151
			Socioeconomic Status	812
			Transfer Status	1,173
	Multiple-Predictor		GISO, RE, SES, TS	790
	Wellbeing Consequences	Stress	FS, GISO	784
		Sadness	FS, GISO	779
		Life Satisfaction	FS, GISO, TS	782
2021	Single-Predictor	Food Security	Gender Identity/Sexuality	780
			Race/Ethnicity	763
			Socioeconomic Status	665
			Transfer Status	755
	Multiple-Predictor		GISO, RE, SES, TS	636
	Wellbeing Consequences	Stress	FS, GISO, RE	736
		Sadness	FS, GISO	773
		Life Satisfaction	FS, GISO, RE, TS	614

Note: FS = Food Security, GIS = Gender Identity/Sexuality, RE = Race/Ethnicity, GSS = Global Socioeconomic Status, FSC = Family Social Class, HSQ = High School Quality, PEL = Parent Education Level, TS = Transfer Status. Table produced using the huxtable package.

Descriptives

Reference Data

Population-level reference data were used in various ways in plots of overall food security estimates and demographics descriptives (below).

For demographics data that come only from admissions data (race/ethnicity and transfer status), population percentages were calculated and presented alongside sample percentages.

For demographics data that exist only in SWaSI data (gender identity/sexuality, global socioeconomic status, family social class, and high school quality), a small, finite population correction was applied to calculation of uncertainty, depicted as error bars representing 95% confidence intervals, around sample estimates of percentages. The correction used the following expression:

$1.96 \times \sqrt{\frac{\hat{p}(1 - \hat{p})}{N_{Sample}} \times \frac{N_{Source} - N_{Sample}}{N_{Source} - 1}} \times 100$

Source N was the total population number of admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2023 (N = 16,778) and spring 2022 (N = 15,754), or admitted, active status, first-through-third-year, U.S. undergraduates during spring 2021 (N = 12,563).

For parent education level, which exists in both admissions data and baseline SWaSI data, population and sample percentages were calculated for admissions data, survey data, and combined data (see this section for combination procedure) and presented alongside each other. The SWaSI baseline survey is not technically population-level, but samples are so large it can be reasonably treated as such for comparison to much smaller end-of-year samples. And in that respect, admissions data for parent education level are not technically population-level either, given substantial missingness, which is why we use a combination of the data sources.

Demographics

Gender Identity/Sexuality

2024

Note: Sample n = 1. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2023. Numerically, the confidence interval was +/- 1.0 percentage points for both 2SLGBTQIA3+ and Not 2SLGBTQIA3+. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

2023

Note: Sample n = 1. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2022. Numerically, the confidence interval was +/- 1.0 percentage points for both 2SLGBTQIA3+ and Not 2SLGBTQIA3+. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

2022

Note: Sample n = 1. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2022. Numerically, the confidence interval was +/- 1.0 percentage points for both 2SLGBTQIA3+ and Not 2SLGBTQIA3+. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

2021

Note: Sample n = 1. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-third-year, U.S. undergraduates during spring 2021. Numerically, the confidence interval was +/- 1.0 percentage points for both 2SLGBTQIA3+ and Not 2SLGBTQIA3+. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

Race/Ethnicity

2024

Note: Population N = 1 (Black/Indigenous was 1% Black), sample n = 1 (Black/Indigenous was 1% Black). Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

2023

Note: Population N = 1 (Black/Indigenous was 1% Black), sample n = 1 (Black/Indigenous was 1% Black). Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

2022

Note: Population N = 1 (Black/Indigenous was 1% Black), sample n = 1 (Black/Indigenous was 1% Black). Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

2021

Note: Population N = 1 (Black/Indigenous was 1% Black), sample n = 1 (Black/Indigenous was 1% Black). Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

Socioeconomic Status

2024

Note: Sample n = 1. When global socioeconomic status was treated as a continuous variable (ranging 1-10) it had a mean of 1.00 and a standard deviation of 1.00. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2023. Numerically, the confidence intervals were +/- 1.0 percentage points for Lower socioeconomic status, +/- 1.0 points for Middle status, and +/- 1.0 points for Upper status. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

2023

Note: Sample n = 1. When global socioeconomic status was treated as a continuous variable (ranging 1-10) it had a mean of 1.00 and a standard deviation of 1.00. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2022. Numerically, the confidence intervals were +/- 1.0 percentage points for Lower socioeconomic status, +/- 1.0 points for Middle status, and +/- 1.0 points for Upper status. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

2022

Note: Sample n = 1. When global socioeconomic status was treated as a continuous variable (ranging 1-10) it had a mean of 1.00 and a standard deviation of 1.00. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-fourth-year, U.S. undergraduates during spring 2022. Numerically, the confidence intervals were +/- 1.0 percentage points for Lower socioeconomic status, +/- 1.0 points for Middle status, and +/- 1.0 points for Upper status. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

2021

Note: Sample n = 1. When global socioeconomic status was treated as a continuous variable (ranging 1-10) it had a mean of 1.00 and a standard deviation of 1.00. Pink error bars represent the 95% confidence interval around the estimated percentage. Calculation of error applied a small, finite population correction using the following expression: 1.96 * sqrt(((proportion * (1 - proportion)) / Sample n) * (Source N - Sample n) / (Source N - 1)) * 100, where Source N in this case was 1 admitted, active status, first-through-third-year, U.S. undergraduates during spring 2021. Numerically, the confidence intervals were +/- 1.0 percentage points for Lower socioeconomic status, +/- 1.0 points for Middle status, and +/- 1.0 points for Upper status. Plot was produced using the ggplot2 package. Colors were produced using the viridis and magma palettes of the viridis package.

Transfer Status

2024

Note: Population N = 1, sample n = 1. Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

2023

Note: Population N = 1, sample n = 1. Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

2022

Note: Population N = 1, sample n = 1. Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

2021

Note: Population N = 1, sample n = 1. Plot was produced using the ggplot2 package. Colors were produced using the viridis palette of the viridis package.

Wellbeing

Distributions

Stress

Note: Violin (colored, wavy) parts of the plot represent the distributions of stress scores by year of administration. Box parts of the plot represent the medians (middle vertical lines in the box) and quartiles (edges of boxs are 1st and 3rd quartiles on the left and right, respectively; lines [“whiskers”] extending outside boxes end at minimum and maximum values on the left and right, respectively) of the distributions by year. Means (M), standard deviations (SD), and sample sizes (n) for each year are numerically depicted along the y-axis. Stress was POMP-scored (Percent of Maximum Possible; J. Cohen et al., 2010). Plot was produced using glue, ggplot2, and ggtext packages. Colors were produced using the viridis palette of the viridis package.

Sadness

Note: Violin (colored, wavy) parts of the plot represent the distributions of sadness scores by year of administration. Box parts of the plot represent the medians (middle vertical lines in the box) and quartiles (edges of boxs are 1st and 3rd quartiles on the left and right, respectively; lines [“whiskers”] extending outside boxes end at minimum and maximum values on the left and right, respectively) of the distributions by year. Means (M), standard deviations (SD), and sample sizes (n) for each year are numerically depicted along the y-axis. Sadness was POMP-scored (Percent of Maximum Possible; J. Cohen et al., 2010). Plot was produced using glue, ggplot2, and ggtext packages. Colors were produced using the viridis palette of the viridis package.

Life Satisfaction

Note: Violin (colored, wavy) parts of the plot represent the distributions of life satisfaction scores by year of administration. Box parts of the plot represent the medians (middle vertical lines in the box) and quartiles (edges of boxs are 1st and 3rd quartiles on the left and right, respectively; lines [“whiskers”] extending outside boxes end at minimum and maximum values on the left and right, respectively) of the distributions by year. Means (M), standard deviations (SD), and sample sizes (n) for each year are numerically depicted along the y-axis. Life satisfaction was POMP-scored (Percent of Maximum Possible; J. Cohen et al., 2010). Plot was produced using glue, ggplot2, and ggtext packages. Colors were produced using the viridis palette of the viridis package.

Statistics

Construct	Year	M	SD	n
Stress	2024	44.6	17.2	1,222
	2023	50.1	18.5	762
	2022	51.1	19.7	1,154
	2021	45.7	18.3	792
Sadness	2024	38.5	24.8	1,211
	2023	46.5	26.8	775
	2022	49.0	27.0	1,148
	2021	45.2	27.5	795
Life Satisfaction	2024	63.0	22.9	1,232
	2023	56.5	23.3	774
	2022	54.2	24.3	1,151
	2021	60.3	23.1	791

Note: M = mean, SD = standard deviation, n = sample size. Stress, sadness, and life satisfaction were POMP-scored (Percent of Maximum Possible; J. Cohen et al., 2010). Table was produced using the huxtable package.

National College Health Assessment

A total of n = 1 first-through-fourth year U.S. undergraduates participated in the National College Health Assessment in spring 2024 (n = 1 had valid data on food security).

Hope Center Student Basic Needs Survey

A total of n = 1 first-through-fourth year U.S. undergraduates participated in the Hope Center Student Basic Needs Survey in winter 2023 (n = 1 had valid data on food security, n = 1 had valid data on housing security, and n = 1 had valid data on homelessness). Hope’s sample was 1.0% 2SLGBTQIA3+, which suggests that SWaSI likely oversampled 2SLGBTQIA3+ students in spring 2023 and that 2022 and 2021 samples are likely more representative with respect to that demographic.

Measurement and Analysis

Food Security

SWaSI

The USDA Food Security Module was implemented in end-of-year check-in waves according to screening methodology.

Implementation

First, students were presented with the following instructions:

page break

Next, students were presented with the first set of items:

If responses to any of the items in the first set were often true or sometimes true, students were presented with the second set of items:

If the response to the above item was “yes,” students were presented with the intermediate followup item below before moving onto the subsequent three items in the second set.

If responses to any of the four yes/no items in the second set was “yes,” students were presented with the following item:

If response to the above item was “yes,” students were presented with the followup item below.

Classification

For each of the 10 items, affirmative responses (i.e., often true, sometimes true, yes, almost every month, several months but not every month) were scored as 1 and negative responses (i.e., never true, no, don’t know, only 1 or 2 months) were scored as 0. Then items were summed, for a possible total score range of 0 to 10. (Students who responded in the negative to the first three items automatically had a total score of 0, because of the screening methodology.) Total score ranges were classified as follows:

0 – High Food Security
1-2 – Marginal Food Security
3-5 – Low Food Security
6-10 – Very Low Food Security

High and Marginal were combined to form the Secure category while Low and Very Low were combined to form the Insecure category.

National College Health Assessment

Implementation

The National College Health Assessment used an adapted version of the short-form of the USDA Food Security Module, which employed a 30-day time frame:

For the following statements, please say whether the statement was often true, sometimes true, or never true for you in the last 30 days.
- The food that I bought just didn’t last, and I didn’t have money to get more (response options: Never True, Sometimes True, Often True)
- I couldn’t afford to eat balanced meals. (response options: Never True, Sometimes True, Often True)
In the last 30 days, did you ever cut the size of your meals or skip meals because there wasn’t enough money for food? (response options: No; Only 1 or 2 days; Yes, some days, but not every day; Yes, almost every day)
In the last 30 days, did you ever eat less than you felt you should because there wasn’t enough money for food? (response options: No, Yes)
In the last 30 days, were you ever hungry but didn’t eat because there wasn’t enough money for food? (response options: No, Yes)

Classification

For each of the 5 items, affirmative responses (i.e., Sometimes True; Often True; Yes, some days, but not every day; Yes, almost every day; Yes) were scored as 1 and negative responses (i.e., Never True, Only 1 or 2 months, No) were scored as 0. Then items were summed, for a possible total score range of 0 to 5. Total score ranges were classified as follows:

0-1 – High/Marginal Food Security
2-4 – Low Food Security
5 – Very Low Food Security

High/Marginal formed the Secure category while Low and Very Low were combined to form the Insecure category.

The National College Health Assessment data were provided with precomputed food security scores and classifications. However, the method is questionable. It appears that Only 1 or 2 days responses were counted as affirmative (i.e., scored as 1) and Yes, some days but not every day and Yes, almost every day responses were counted twice (i.e., scored as 2), perhaps because it is a combination of a yes/no question and a frequency question from the original module. Both these alterations substantially change the measurement. The second (doubling) alteration may be legitimate, but the 1st (Only 1 or 2 days as affirmative) is inconsistent with USDA scoring guidelines. Without psychometric validation of either alteration, this scoring procedure seems unjustifiable. It almost certainly overestimates food insecurity. Hence, we computed a 0-5 sum score based on the criteria described at the beginning of this subsection and kept the same classification ranges (albeit scores of 6 are not part of the Very Low Food Security range). This scoring is likely to, at worst, underestimate food insecurity and, at best, estimate it accurately.

Hope Center Student Basic Needs Survey

Implementation

Like SWaSI, Hope used the 10-item USDA Food Security Module. The differences in implementation between Hope and SWaSI were:

Hope used a 30-day time frame rather than 12-month.
Hope did not include response options in the first set of items (i.e., the first set of items did not include “Was that often true, sometimes true, or never true for you in the last 12 months?” after the statement to which people were responding). This is a very reasonable adaptation we may adopt in future measurement through SWaSI.
Hope’s phrasing of the “how often” follow-up questions was different (e.g., “In the last 30 days, how often did this happen (you cut the size of your meals or skipped meals because there wasn’t enough money for food)?” rather than “How often did you cut the size of your meals or skip meals because there wasn’t enough money for food?”). We prefer our adaptation to theirs.
Response options for the “how often” follow-up questions were also different (Once, Twice, Three times, Four times, Five times, More than five times). This difference is largely a function of recall time frame.
An additional set of food security items were presented to students who are parents (see below for items).

Classification

Because of their extra parent-related items, Hope scored food security items differently depending on whether a student did or did not have children.

For students with no children, and consistent with SWaSI scoring, possible total scores ranged from 0 to 10 and each affirmative response counted as 1 point. Students with children answered a set of questions that pertained to feeding their children. Each affirmative response to the regular 10-item module counted as 1 point and affirmative responses to each of the following parenting-relevant questions got an additional point:

I relied on only a few kinds of low-cost food to feed my child/children because I was running out of money to buy food.
I couldn’t feed my child/children a balanced meal because I couldn’t afford that.
My child/children were not eating enough because I just couldn’t afford enough food.
In the last 30 days, did you ever cut the size of your child/children’s meals because there wasn’t enough money for food?
In the last 30 days, were your child/children ever hungry but you just couldn’t afford more food?
In the last 30 days, did your child/children ever skip a meal because there wasn’t enough money for food?
In the last 30 days, did your child/children ever not eat for a whole day because there wasn’t enough money for food?
In the last 30 days, how often did your child/children skip meals due to lack of money or food?

Total score ranges were classified as follows:

Students with no children
- 0 – High Food Security
- 1-2 – Marginal Food Security
- 3-5 – Low Food Security
- 6-10 – Very Low Food Security
Students with children
- 0 – High Food Security
- 1-2 – Marginal Food Security
- 3-7 – Low Food Security
- 8-18 – Very Low Food Security

High and Marginal were combined to form the Secure category while Low and Very Low were combined to form the Insecure category.

Demographics

Gender identity/sexuality was derived from two items: gender identity and sexual orientation. Data are collected at the end of both baseline and end-of-year check-in waves of SWaSI. Data from baseline were used to fill missing data from end-of-year waves.

For gender identity, students were asked “What is your gender identity? (select all that apply).” Response options changed in baseline for the 2020-21 cohort and again in end-of-year check-in in spring 2023. At baseline, among the 2018-19 and 2019-20 cohorts, response options were Fluid, Gender Queer, Man, Transgender, Woman, None of these describe me well; this is better:, and Prefer not to respond, in that order from top to bottom. At baseline, among the 2020-21, 2021-22, and 2022-23 cohorts, options were Agender, Genderqueer, Man, Nonbinary (including gender fluid, gender nonconforming, etc.), Transgender, Trans Man, Trans Woman, Woman, Questioning or unsure, None of these describe me well; this is better:, and prefer not to respond, in that order from top to bottom. In both versions, the None of these describe me well; this is better: and prefer not to respond options were mutually exclusive to the other options. At end-of-year check-in in spring 2023, response options were Agender, Cisgender, Genderfluid, Gender nonconforming, Genderqueer, Man, Multigender (including bigender, polygender, pangender, etc.), Nonbinary, Third Gender, Transgender, Two-Spirit, Woman, Questioning or exploring, Not listed (please specify):, and prefer not to respond, in that order from top to bottom. In all versions, the None of these describe me well; this is better: or Not listed (please specify): option included an open text box for students to specify how they identified, most of which were coded as one or more of the existing categories. The small few that were truly unique were excluded from further categorization and analysis.

For sexual orientation, incoming students were asked “Which term best describes your sexual orientation?” The question changed from single-select to multi-select in baseline for the 2020-21 cohort, adding “(select all that apply)” to the item and response options changed in baseline for the 2020-21 cohort. Response options changed again in end-of-year check-in in spring 2023. At baseline, among the 2018-19 and 2019-20 cohorts, response options were Asexual, Bisexual, Gay or Lesbian, Queer, Pansexual, Straight or heterosexual, None of these describe me well; this is better:, and prefer not to respond, in that order from top to bottom. At baseline, among the 2020-21, 2021-22, and 2022-23 cohorts, options were Asexual, Bisexual, Gay, Heterosexual or straight, Lesbian, Pansexual, Queer, Questioning or unsure, Same-gender loving, None of these describe me well; this is better:, and prefer not to respond, in that order from top to bottom. In both versions, the None of these describe me well; this is better: and prefer not to respond options were mutually exclusive to the other options. At end-of-year check-in in spring 2023, response options were Asexual, Bisexual, Fluid or flexible, Gay, Graysexual, Heterosexual or straight, Lesbian, Pansexual, Queer, Questioning or exploring, Same-gender loving, Not listed (please specify):, and prefer not to respond, in that order from top to bottom. In all versions, the None of these describe me well; this is better: or Not listed (please specify): option included an open text box for students to specify how they identified, most of which were coded as one or more of the existing categories. The small few that were truly unique were thus excluded from further categorization and analysis.

All categories of gender identity that were not exclusively Man or Woman were grouped together to form a Nonbinary category. All categories of sexuality that were not exclusively Straight or heterosexual/Heterosexual or straight were grouped together to form an LGBQ+ category. Then, the Nonbinary and LGBQ+ categories were grouped together to form 2SLGBTQIA3+. In analyses, Not 2SLGBTQIA3+ was coded as 0 and 2SLGBTQIA3+ was coded as 1.

Race/ethnicity data were obtained from admissions records. Prospective students were first asked to identify whether they are Hispanic or Latino (yes, no) and then were asked to identify racial categories that apply to them (American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, White). If students selected “no” regarding ethnicity (i.e., whether they are Hispanic or Latino) and selected more than one racial category, they were categorized as “Two or more races,” which corresponds to the Multiracial/ethnic category in this document. If they selected “yes” regarding ethnicity and selected at least one racial category, they were nonetheless categorized as Hispanic or Latino. To be more contemporary, inclusive, and efficient we relabeled American Indian or Alaska Native as Native American/Alaska Native, Black or African American as Black, Hispanic or Latino as Latine/a/o/x, and Native Hawaiian or Other Pacific Islander as Native Hawaiian/Pacific Islander. The Black, Native American/Alaska Native, and Native Hawaiian/Pacific Islander categories were grouped together to form the Black/Indigenous category because there were too few participants in each of those groups to be analyzed on their own. In analyses, categories were weighted effects coded (J. Cohen et al., 2003) with White as the base category.

Global socioeconomic status data were collected at the end of the baseline wave of SWaSI, where incoming students were asked:

Note: See Goodman et al. (2001) for source of measurement. In analyses, responses were coded 1 through 10 corresponding to rungs on the ladder. In depicting results, the scale was condensed to three ordinal categories: Lower (1 through 4, which was approximately -1SD), Middle (5 through 7), and Upper (8, which was approximately +1SD, through 10).

Family social class data were collected at the end of the baseline wave of SWaSI, where incoming students were asked “How would you describe your family’s social class?” Response options were Working Class, Lower-Middle Class, Middle Class, Upper-Middle Class, and Upper Class, in that order from top to bottom. In analyses, responses were coded 1 through 5, from working to upper. In depicting results, the scale was condensed to three ordinal categories: Lower (Working Class and Lower-Middle Class), Middle (Middle Class), and Upper (Upper-Middle Class, and Upper Class).

High school quality data were collected at the end of the baseline wave of SWaSI, where incoming students were asked “How do you think the high school you attended compares to the high schools attended by most incoming UO students?” Response options were My high school is less advantaged than the high schools attended by most other incoming students, My high school is neither less nor more advantaged than the high schools attended by most other incoming students, and My high school is more advantaged than the high schools attended by most other incoming students, in that order from top to bottom. In analyses, responses were weighted effects coded (J. Cohen et al., 2003) with “neither” as the base category.

Parent education level data were a combination of admissions data and data collected at the end of the baseline wave of SWaSI. Admissions applications ask prospective students what their parents’ education levels are (from the Common Application: Some grade/primary school, Completed grade/primary school, Some high school/secondary school, Graduated from high/secondary school [or equivalent], Some trade school or community college, Some college/university, Graduated from college/university, Graduate school). At the end of the baseline wave of SWaSI, students were asked:

Students who had at least one parent who “Graduated from college/university” or has a “4-year college / university degree” were considered Continuing-Generation and students who did not have any parents who “Graduated from college/university” or had a “4-year college / university degree” were considered First-Generation. Data from the baseline wave of SWaSI were used to fill data missing from admissions records. In analyses, Continuing-Generation was coded as 0 and First-Generation was coded as 1.

Transfer status data were obtained from admissions records. Students who were coming from another institution of higher education were considered Transfer and students who were coming from high school were considered First-Time. Transfer students were included in data collection starting with the 2018-19 cohort, which is the beginning of the year-in-school time frame for any analysis presented here. In analyses, First-Time was coded as 0 and Transfer was coded as 1.

Wellbeing Constructs

Stress refers to the extent to which students feel able to control things in life, including handling difficulties as they arise (S. Cohen et al., 1983). It is measured with the items below using the following response coding/options: 1 = never, 2 = almost never, 3 = sometimes, 4 = fairly often, 5 = very often. Items were POMP-scored (J. Cohen et al., 2010) to form a composite variable prior to analysis.

In the last month, how often have you felt that you were unable to control the important things in your life?
In the last month, how often have you felt confident about your ability to handle your personal problems?
In the last month, how often have you felt that things were going your way?
In the last month, how often have you felt difficulties were piling up so high that you could not overcome them?

Sadness refers to a cluster of students’ feelings centered on emotional suffering and social isolation (Watson & Clark, 1999). It is measured with the prompt “Indicate to what extent you have felt this way during the past month.” and five emotion descriptors (sad, blue, downhearted, alone, lonely) as items using the following response coding/options: 1 = very slightly or not at all, 2 = a little, 3 = moderately, 4 = quite a bit, 5 = extremely. Items were POMP-scored (J. Cohen et al., 2010) to form a composite variable prior to analysis.

Life Satisfaction refers to students’ global judgments of how satisfactory their lives are in general (Deiner et al., 1985). It is measured with the items below using the following response coding/options: 1 = strongly disagree, 2 = moderately disagree, 3 = slightly disagree, 4 = slightly agree, 5 = moderately agree, 6 = strongly agree. Items were POMP-scored (J. Cohen et al., 2010) to form a composite variable prior to analysis.

In most ways, my life is close to my ideal.
The conditions of my life are excellent.
I am satisfied with my life.
So far, I have gotten the important things I want in life.
If I could live my life over, I would change almost nothing.

Basic Needs Support

Implementation

Starting in 2022, at the end of the end-of-year check-in survey, students were asked: How can the UO better support your basic needs (food, housing, safety, belonging, self-esteem, fulfillment)? They were provided with a essay text box to write their responses.

Analysis

Rows with no response at all were removed (n₂₀₂₃ = 1; n₂₀₂₂ = 1). Then, responses that were effectively nonresponses (e.g., “I don’t know,” “N/A,” “none,” “nothing”) were removed (n₂₀₂₃ = 1; n₂₁₂₂ = 1). Finally, to focus on what could be improved and how, responses which indicated that students felt their basic needs were being supported well by UO, that their needs were otherwise satisfied, or felt that their own need satisfaction was their personal responsibility and not the university’s were removed (n₂₀₂₃ = 1; n₂₁₂₂ = 1). The final number of responses used in analyses were: n₂₀₂₃ = 1; n₂₁₂₂ = 1. Prior to subsequent analyses, responses were converted to all lower case, punctuation was removed, and contractions were uncontracted (e.g., “i’ve” to “i have”), except words ending in “apostrophe s” (i.e., ’s) because they have no straightforward uncontraction, as they could be some sort of contraction or could be possessive. Instances of ’s were deleted.

Hierarchy of Needs

The str_detect function of the stringr package, which is part of the greater tidyverse, was used to identify word mentions, which were then counted to represent how many students wrote about need domains and financial, pay/wage, and unionization issues. Percentages were computed by dividing those numbers by the total number of usable responses and multiplying by 100.

The words “food” and “housing” represented physiological needs, some version of the word “safe” represented safety needs, some version of the word “belong” represented belonging needs, the word “esteem” represented esteem needs, and some version of the word “fulfill” represented fulfillment (or self-actualization) needs. A list of financially-related words was developed (“afford,” “bills,” “bucks,” “buy,” “bought,” “cash,” “cheap,” “cost,” “debt,” “dollars,” “econom,” “expens,” “fee,” “financ,” “fund,” “gentrif,” “scholarship,” “spendy,” “loan,” “money,” “overpriced,” “paid,” “pay,” “poverty,” “price,” “stipend,” “subsidi,” “subsidy,” “tuition,” “tution[sic],” “wage”). The words “pay,” “paid,” and “wage” were specifically targeted within that list. And the word “union” was searched in 2023 data (unionization efforts were irrelevant prior to this wave). Instances that referred to the student union (i.e., the Erb Memorial Union) were subsequently discounted.

All of the above involved counting how many students mentioned words. We also counted how many times words were mentioned by first using the tm package to make a document-term matrix in which numbers and stop words were first removed and then words with no or very little descriptive utility were subsequently removed to allow more descriptively useful words to the top. Specifically, the following words were removed: also

Content Analysis

Themes

Responses were open-coded using Grounded Theory (Strauss & Corbin, 2006; Tracy, 2019), which is a framework used to understand people’s experiences by building on themes from qualitative data they provide.

Wordnet

The udpipe package was used to annotate parts of speech prior to analysis. Coocurrences of nouns and adjectives (skipgram = 1) were computed using the cooccurrence function of the udpipe package. Coocurrences were visualized using the igraph and ggraph packages, imposing a minimum coocurrence threshold of 3.

Wordcloud

The udpipe package was used to annotate parts of speech prior to analysis and to extract keywords using the Rapid Keyword Extraction (RAKE) algorithm (Rose et al., 2010). Keywords were visualized using the wordcloud2 package, imposing a minimum occurrence threshold of 3.

Quotes

Direct quotes were selected (and silently edited for clarity) to represent take-home messages represented in themes, wordnets, and wordclouds.

Housing Insecurity and Homelessness

The Hope Center Student Basic Needs Survey measures housing insecurity and homelessness using the following methods.

Implementation

First, students were presented with the following instruction:

Next, we have some questions about what your life is like outside of school and work.

Then, students were presented with several items about their living situation. Items relevant to housing insecurity and homelessness indices are presented below.

Which of the following have you experienced in the past 12 months? (Please check all that apply.)

Been unable to pay or underpaid rent or mortgage (1pt)
Been evicted from your home
Received a summons to appear in housing court (1pt)
Not paid the full amount for utilities (such as gas, oil, electric, water, internet, phone) (1pt)
Borrowed money from friends or family to help pay bills
Had an account default or go into collections (1pt)
Moved in with other people, even for a little while, because of financial problems (1pt)
Lived with others beyond the expected capacity of the house or apartment (1pt)
Had trouble finding an affordable place to live on or near campus
None of the above

In the past 12 months, was there a rent or mortgage increase that made it difficult to pay?

Yes (1pt)
No

In the past 12 months, how many times have you moved (including for college)?

None
Once
Twice
Three times (1pt)
Four times (1pt)
Five or more times (1pt)

In the past 12 months, have you LEFT your household because you felt unsafe?

Yes (1pt)
No

In the past 12 months, have you ever been homeless?

Yes (1pt)
No

In the past 12 months have you slept in any of the following places? (Please check all that apply.)

Campus or university housing
Sorority/fraternity house
In a rented or owned house, mobile home, or apartment (alone or with roommates/friends)
In a rented or owned house, mobile home, or apartment (with family)
Temporarily stayed with relatives, friends, or couch surfing until I found other housing (1pt)
Temporarily at a hotel or motel without a permanent home to return to (not on vacation or business travel) (1pt)
At a shelter (1pt)
In transitional housing or independent living program (1pt)
At a group home such as a halfway house or residential program for mental health or substance abuse (1pt)
At a treatment center (such as detox, hospital, etc.) (1pt)
In a camper or RV (not on vacation) (1pt)
An outdoor location such as street, sidewalk, or alley, bus or train stop, campground or woods, park, beach, or riverbed, under bridge or overpass (1pt)
In a closed area/space with a roof not meant for human habitation such as an abandoned building, car or truck, van, encampment or tent, or unconverted garage, attic, or basement (1pt)

Classification

Each affirmative response counted as 1 point. Possible total scores ranged from 0 to 9 for the housing insecurity index and from 0 to 10 for the homelessness index. Response options highlighted in blue and marked with (1pt) were part of the housing insecurity index and those highlighted in teal and marked with (1pt) were part of the homelessness index. Total scores on housing insecurity were classified as 0 = Housing Secure and greater than 0 = Housing Insecure. Total scores on homelessness were classified as 0 = Not Experiencing Homelessness and greater than 0 = Experiencing Homelessness.

Student Wellbeing and Success Initiative

Undergraduate Food Insecurity 2021-2024

Brian A. M. Clark, Ph.D.

Renée Delgado-Riley, Ph.D.

Amy N. Warnock, M.S.

What is This Document for?

Who is SWaSI?

What is Food Security?

Brief Method

Summary of Findings

Recommendations

Overall Estimates

2024

2023

2022

2021

By Demographics

Zero-Order Relationships

2024

Disability/Neurodivergence

Gender Identity/Sexuality

Race/Ethnicity

Socioeconomic Status

Transfer Status

2023

Disability/Neurodivergence

Gender Identity/Sexuality

Race/Ethnicity

Socioeconomic Status

Transfer Status

2022

Disability/Neurodivergence

Gender Identity/Sexuality

Race/Ethnicity

Socioeconomic Status

Transfer Status

2021

Disability/Neurodivergence

Gender Identity/Sexuality

Race/Ethnicity

Socioeconomic Status

Transfer Status

Accounting for Unique Variation

2024

2023

2022

2021

Wellbeing Consequences

2024

Stress

Sadness

Life Satisfaction

2023

Stress

Sadness

Life Satisfaction

2022

Stress

Sadness

Life Satisfaction

2021

Stress

Sadness

Life Satisfaction

Student Voices – Basic Needs Support

Results

2024

Hierarchy of Needs

Content Analysis

Themes

Overview

Frequencies

Wordnet

Wordcloud

Quotes

2023

Hierarchy of Needs

Content Analysis

Themes

Overview